Finite Group Theory

I. Martin Isaacs

Finite Group Theory

I. Martin Isaacs

Graduate Studies in Mathematics

Volume 92

American Mathematical Society Providence, Rhode Island

Editorial Board

David Cox (Chair) Steven G. Krantz Rafe Mazzeo Martin Scharlemann

2000 M a t h e m a t i c s S u b j e c t C l a s s i f i c a t i o n . P r i m a r y 2 0 B 1 5 , 2 0 B 2 0 , 2 0 D 0 6 , 2 0 D 1 0 , 2 0 D 1 5 , 20D20, 20D25, 20D35, 20D45, 20E22, 20E36.

For

additional information a n d updates o n this book, visit www.ams.org/bookpages/gsm-92

Library of Congress Cataloging-in-Publication Data Isaacs, I. Martin, 1940Finite group theory / I. Martin Isaacs. p. cm. (Graduate studies in mathematics ; v. 92) Includes index. ISBN 978-0-8218-4344-4 (alk. paper) 1. Finite groups. 2. Group theory. I. Title. QA177.I835 2008 512'.23dc22

2008011388

Copying and reprinting. Individual readers of this publication, and nonprofit libraries acting for them, are permitted to make fair use of the material, such as to copy a chapter for use in teaching or research. Permission is granted to quote brief passages from this publication in reviews, provided the customary acknowledgment of the source is given. Republication, systematic copying, or multiple reproduction of any material in this publication is permitted only under license from the American Mathematical Society. Requests for such permission should be addressed to the Acquisitions Department, American Mathematical Society, 201 Charles Street, Providence, Rhode Island 02904-2294, USA. Requests can also be made by e-mail to reprint-permissionaams.org. 2008 by the American Mathematical Society. All rights reserved. The American Mathematical Society retains all rights except those granted to the United States Government. Printed in the United States of America. The paper used in this book is acid-free and falls within the guidelines established to ensure permanence and durability. Visit the AMS home page at http://www.ams.org/ 10 9 8 7 6 5 4 3 2 1 13 12 11 10 09 08

To

Deborah

Contents

Preface C h a p t e r 1. C h a p t e r 2. C h a p t e r 3. C h a p t e r 4. C h a p t e r 5. C h a p t e r 6. C h a p t e r 7. C h a p t e r 8. C h a p t e r 9. Sylow T h e o r y Subnormality Split Extensions Commutators Transfer Frobenius A c t i o n s T h e T h o m p s o n Subgroup Permutation Groups M o r e on S u b n o r m a l i t y

ix 1 45 65 113 147 177 201 223 271 295 325 345

C h a p t e r 10. M o r e Transfer T h e o r y Appendix: Index T h e Basics

Preface

T h i s b o o k is a somewhat expanded version of a graduate course i n finite group theory that I often teach at the U n i v e r s i t y of W i s c o n s i n . I offer this course i n order to share what I consider to be a beautiful subject w i t h as m a n y people as possible, and also to provide the solid b a c k g r o u n d i n pure group theory t h a t m y d o c t o r a l students need to carry out their thesis work i n representation theory. T h e focus of group theory research has changed profoundly i n recent decades. S t a r t i n g near the beginning of the 20th century w i t h the work of W . B u r n s i d e , the major p r o b l e m was to find a n d classify the finite simple groups, and indeed, m a n y of the most significant results i n pure group theory and i n representation theory were directly, or at least peripherally, related to this goal. T h e simple-group classification now appears to be complete, and current research has shifted to other aspects of finite group theory i n c l u d i n g p e r m u t a t i o n groups, p-groups and especially, representation theory. It is certainly no less essential i n this post-classification p e r i o d that group-theory researchers, whatever their subspecialty, s h o u l d have a mastery of the classical techniques and results, and so w i t h o u t a t t e m p t i n g to be encyclopedic, I have included m u c h of t h a t m a t e r i a l here. B u t m y choice of topics was largely determined by m y p r i m a r y goal i n w r i t i n g this book, w h i c h was to convey to readers m y feeling for the beauty and elegance of finite group theory. G i v e n its origin, this b o o k should certainly be suitable as a text for a graduate course like mine. B u t I have t r i e d to write it so that readers w o u l d also be comfortable using it for independent study, a n d for t h a t reason, I have t r i e d to preserve some of the informal flavor of m y classroom. I have tried to keep the proofs as short and clean as possible, b u t w i t h o u t o m i t t i n g

ix

Preface

details, and indeed, i n some of the more difficult m a t e r i a l , m y arguments are simpler t h a n can be found i n p r i n t elsewhere. F i n a l l y , since I firmly believe t h a t one cannot learn mathematics w i t h o u t doing it, I have included a large number of problems, m a n y of w h i c h are far from routine. Some of the m a t e r i a l here has rarely, if ever, appeared previously i n books. Just i n the first few chapters, for example, we offer Zenkov's marvelous theorem about intersections of abelian subgroups, W i e l a n d t ' s "zipper l e m m a " i n s u b n o r m a l i t y theory and a proof of Horosevskii's theorem that the order of a group a u t o m o r p h i s m can never exceed the order of the group. L a t e r chapters include m a n y more advanced topics that are h a r d or impossible to find elsewhere. M o s t of the students who attend m y group-theory course are second-year graduate students, w i t h a substantial m i n o r i t y of first-year students, and an occasional well-prepared undergraduate. A l m o s t all of these people had previously been exposed to a standard first-year graduate abstract algebra course covering the basics of groups, rings and fields. I expect that most readers of this b o o k w i l l have a similar background, and so I have decided not to begin at the beginning. M o s t of m y readers (like m y students) w i l l have previously seen basic group theory, so I wanted to avoid repeating that m a t e r i a l and to start w i t h something more exciting: Sylow theory. B u t I recognize that m y audience is not homogeneous, and some readers w i l l have gaps i n their preparation, so I have included an a p p e n d i x that contains most of the assumed m a t e r i a l in a fairly condensed form. O n the other hand, I expect that m a n y i n m y audience w i l l already know the Sylow theorems, but I a m confident that even these well-prepared readers w i l l find m a t e r i a l that is new to t h e m w i t h i n the first few sections. M y semester-long graduate course at W i s c o n s i n covers most of the first seven chapters of this book, starting w i t h the Sylow theorems and culm i n a t i n g w i t h a purely group-theoretic proof of B u r n s i d e ' s famous p q theorem. Some of the topics along the way are s u b n o r m a l i t y theory, the Schur-Zassenhaus theorem, transfer theory, coprime group actions, Frobenius groups, and the n o r m a l p-complement theorems of Frobenius and of T h o m p s o n . T h e last three chapters cover material for w h i c h I never have time i n class. C h a p t e r 8 includes a proof of the s i m p l i c i t y of the groups P S L ( n , q ) , and also some graph-theoretic techniques for s t u d y i n g subdegrees of p r i m i t i v e and n o n p r i m i t i v e p e r m u t a t i o n groups. S u b n o r m a l i t y theory is revisited i n C h a p t e r 9, w h i c h includes W i e l a n d t ' s beautiful a u t o m o r p h i s m tower theorem and the T h o m p s o n - W i e l a n d t theorem related to the Sims
a b

Preface

xi

conjecture. F i n a l l y , C h a p t e r 10 presents some advanced topics i n transfer theory, i n c l u d i n g Y o s h i d a ' s theorem a n d the so-called "principal ideal theorem". F i n a l l y , I t h a n k m y m a n y students and colleagues who have contributed ideas, suggestions and corrections while this b o o k was being w r i t t e n . I n particular, I m e n t i o n that the comments of Y a k o v B e r k o v i c h and G a b r i e l N a v a r r o were invaluable and very m u c h appreciated.

Chapter 1

Sylow Theory

1A It seems appropriate to begin this b o o k w i t h a t o p i c t h a t underlies v i r t u a l l y all of finite group theory: the Sylow theorems. I n this chapter, we state and prove these theorems, a n d we present some applications a n d related results. A l t h o u g h m u c h of this m a t e r i a l should be very familiar, we suspect t h a t most readers w i l l find t h a t at least some of the content of this chapter is new to t h e m . A l t h o u g h the theorem t h a t proves Sylow subgroups always exist dates back to 1872, the existence proof t h a t we have decided to present is t h a t of H . W i e l a n d t , published i n 1959. W i e l a n d t ' s proof is slick a n d short, but it does have some drawbacks. It is based o n a t r i c k t h a t seems to have no other a p p l i c a t i o n , a n d the proof is not really constructive; it gives no guidance about how, i n practice, one might a c t u a l l y find a Sylow subgroup. B u t W i e l a n d t ' s proof is beautiful, a n d t h a t is the p r i n c i p a l m o t i v a t i o n for presenting it here. A l s o , W i e l a n d t ' s proof gives us an excuse to present a quick review of the theory of group actions, w h i c h are nearly as u b i q u i t o u s i n the study of finite groups as are the Sylow theorems themselves. W e devote the rest of this section to the relevant definitions a n d basic facts about actions, a l t h o u g h we o m i t some details from the proofs. Let G be a group, and let ft be a n o n e m p t y set. (We w i l l often refer to the elements of ft as "points".) Suppose we have a rule t h a t determines a new element of ft, denoted a-g, whenever we are given a point a e ft and a n element g e G. W e say t h a t this rule defines a n a c t i o n of G on ft i f the following two conditions hold.

1. S y l o w

Theory

(1) a - l = a for a l l a G ft and (2) { a - g ) - h = a-fo/i) for a l l a G ft and a l l group elements g , h e G .

Suppose t h a t G acts on ft. It is easy to see t h a t i f g G G is arbitrary, then the function a : ft -> ft defined by - a-$ has an inverse: the function C T _ I . Therefore, a is a p e r m u t a t i o n of the set ft, w h i c h means t h a t o is b o t h injective and surjective, and thus a lies i n the s y m m e t r i c group Sym(ft) consisting of a l l permutations of ft. In fact, the m a p g -> a is easily seen to be a h o m o m o r p h i s m from G into Sym(ft). ( A h o m o m o r p h i s m like this, w h i c h arises from an action of a group G on some set, is called a p e r m u t a t i o n representation of G.) T h e kernel of this h o m o m o r p h i s m is, of course, a n o r m a l subgroup of G, w h i c h is referred to as the kernel of the action. T h e kernel is exactly the set of elements g e G t h a t act t r i v i a l l y on ft, w h i c h means t h a t a - g = a for a l l points a e ft.
g
5

Generally, we consider a theorem or a technique t h a t has the power to find a n o r m a l subgroup of G to be "good", and indeed p e r m u t a t i o n representations can be good i n this sense. (See the problems at the end of this section.) B u t our goal i n i n t r o d u c i n g group actions here is not to find n o r m a l subgroups; it is to count things. Before we proceed i n that direction, however, it seems appropriate to mention a few examples. Let G be arbitrary, and take ft = G. W e can let G act on G by right m u l t i p l i c a t i o n , so t h a t x - g = x g for x , g G G. T h i s is the regular action of G, a n d it should be clear t h a t it is faithful, w h i c h means t h a t its kernel is t r i v i a l . It follows t h a t the corresponding p e r m u t a t i o n representation of G is an i s o m o r p h i s m of G into S y m ( G ) , and this proves C a y l e y ' s theorem: every group is isomorphic to a group of permutations on some set. We continue to take ft = G, but this time, we define x - g = g ~ x g . (The standard n o t a t i o n for g ~ x g is x.) It is t r i v i a l to check t h a t x = x and that { x 9 ) = 9 for a l l x , g , h G G, and thus we t r u l y have a n action, w h i c h is called the conjugation action of G on itself. N o t e t h a t x = x i f and o n l y i f x g = gx, and thus the kernel of the conjugation action is the set of elements g e G t h a t commute w i t h a l l elements x e G. T h e kernel, therefore, is the center Z ( G ) .
x l h h X 9 l

A g a i n let G be arbitrary. I n each of the previous examples, we took ft = G , but we also get interesting actions i f instead we take ft to be the set of a l l subsets of G. I n the conjugation action of G on ft we let X - g = X = { x \ x e X } and i n the r i g h t - m u l t i p l i c a t i o n action we define X - g = X g = { x g | x 6 X } . O f course, i n order to make these examples work, we do not really need ft to be a l l subsets of G. F o r example, since a conjugate of a subgroup is always a subgroup, the conjugation action is well defined if we take ft to be the set of a l l subgroups of G. A l s o , b o t h right m u l t i p l i c a t i o n
9 9

1A

a n d conjugation preserve cardinality, and so each of these actions makes sense i f we take ft to be the collection of a l l subsets of G of some fixed size. In fact, as we shall see, the t r i c k i n W i e l a n d t ' s proof of the Sylow existence theorem is to use the right m u l t i p l i c a t i o n action of G on its set of subsets w i t h a certain fixed cardinality. We mention one other example, w h i c h is a special case of the rightm u l t i p l i c a t i o n action on subsets that we discussed i n the previous paragraph. Let H C G be a subgroup, and let ft = { H x \ x G G } , the set of right cosets of H i n G. If X is any right coset of H , it is easy to see that X g is also a right coset of H . (Indeed, if X = H x , then X g = H { x g ) . ) T h e n G acts on the set ft by right m u l t i p l i c a t i o n . In general, i f a group G acts on some set ft a n d a G ft, we write G = { g G G | a - g = a } . It is easy to check t h a t G is a subgroup of G ; it is called the stabilizer of the point a . For example, i n the regular action of G on itself, the stabilizer of every point (element of G ) is the t r i v i a l subgroup. In the conjugation action of G on G, the stabilizer of x G G is the centralizer C ( x ) and i n the conjugation a c t i o n of G on subsets, the stabilizer of a subset X is the normalizer N ( X ) . A useful general fact about point stabilizers is the following, w h i c h is easy to prove. In any action, if a - g = /?, then the stabilizers G a n d G p are conjugate i n G, and i n fact, { G ) 3 = G.
a a G G a a p

Now consider the action (by right m u l t i p l i c a t i o n ) of G on the right cosets of H , where H C G is a subgroup. T h e stabilizer of the coset H x is the set of a l l group elements g such that H x g = H x . It is easy to see t h a t g satisfies this c o n d i t i o n if and only if x g G H x . ( T h i s is because two cosets H u and H v are identical if and only if u G H v . ) It follows t h a t g stabilizes H x if and only if g G x ~ H x . Since x ~ H x = H , we see that the stabilizer of the point (coset) H x is exactly the subgroup H , conjugate to H v i a x . It follows t h a t the kernel of the action of G on the right cosets of H i n
l l x x

G is exactly
G

f| H
xeG

. T h i s subgroup is called the core of H i n G, denoted

c o r e ( H ) . T h e core of H is n o r m a l i n G because i t is the kernel of an action, and, clearly, it is contained i n H . In fact, if AT < G is any n o r m a l subgroup that happens to be contained i n H , then N = N C H for a l l x G G , and thus N C c o r e ( t f ) . In other words, the core of H i n G is the unique largest n o r m a l subgroup of G contained i n H . (It is "largest" i n the strong sense t h a t it contains a l l others.)
x x G

We have digressed from our goal, w h i c h is to actions to count things. B u t having come this far, results that our discussion has essentially proved. theorem and its corollaries can be used to prove subgroups, and so they might be considered to be

show how to use group we m a y as well state the N o t e t h a t the following the existence of n o r m a l "good" results.

1. S y l o w

Theory

1.1. T h e o r e m . L e t H C G be o f H i n G. T h e n G / c o v e { H ) p a r t i c u l a r , if t h e i n d e x \ G : H s u b g r o u p of S , t h e s y m m e t r i c
G n

a s u b g r o u p , a n d l e t ft be t h e set of r i g h t cosets i s i s o m o r p h i c t o a s u b g r o u p o/Sym(ft). I n \ = n , then G / c o r e ( H ) is isomorphic to a group o nn symbols.

G

P r o o f . T h e action of G on the set ft by right m u l t i p l i c a t i o n defines a h o m o m o r p h i s m 6 (the p e r m u t a t i o n representation) from G into Sym(ft). Since ker(0) = c o r e ( t f ) , it follows by the h o m o m o r p h i s m theorem t h a t G / c o r e { H ) ^ 0 ( G ) , w h i c h is a subgroup of S y m ( G ) . T h e last statement follows since if \ G : H \ = n , then (by definition of the index) |ft| = n , and thus Sym(ft) = S . U
G G n

1.2. C o r o l l a r y . L e t G be a g r o u p , a n d suppose that H C G is a subgroup w i t h \ G : H \ = n . T h e n H c o n t a i n s a n o r m a l s u b g r o u p N of G such that \ G : N \ divides n\. P r o o f . Take N = c o r e ( H ) . T h e n G / N is isomorphic to a subgroup of the s y m m e t r i c group S , and so by Lagrange's theorem, \ G / N \ divides \S \=n\.
G n n

1.3. C o r o l l a r y . L e t G be s i m p l e a n d c o n t a i n a s u b g r o u p of i n d e x n > 1. T h e n \G\ divides n l . P r o o f . T h e n o r m a l subgroup N of the previous corollary is contained i n H , a n d hence it is proper i n G because n > 1. Since G is simple, N = 1 , a n d thus \ G \ = \ G / N \ divides n l . U In order to pursue our m a i n goal, w h i c h is counting, we need to discuss the "orbits" of an action. Suppose t h a t G acts on ft, a n d let a G ft. T h e set O = { a - g | g G G } is called the orbit of a under the given action. It is routine to check t h a t i f 3 G O , then O p = O , and it follows that distinct orbits are a c t u a l l y disjoint. A l s o , since every point is i n at least one orbit, it follows t h a t the orbits of the action of G on ft p a r t i t i o n ft. I n particular, if ft is finite, we see t h a t |ft| = where i n this s u m , O runs over the
a a a

full set of G - o r b i t s on ft. We mention some examples of orbits a n d orbit decompositions. F i r s t , i f H C G is a subgroup, we can let H act on G by right m u l t i p l i c a t i o n . It is easy to see t h a t the orbits of this action are exactly the left cosets of H i n G . (We leave to the reader the p r o b l e m of realizing the right cosets of H in G as the orbits of an appropriate action of H . B u t be careful: the rule x - h - hx does n o t define an action.) Perhaps it is more interesting to consider the conjugation action of G on itself, where the orbits are exactly the conjugacy classes of G . T h e fact

1A

t h a t for a finite group, the order \ G \ is the s u m of the sizes of the classes is sometimes called the class equation of G. How 1.4. b i g is an orbit? T h e key result here is the following. (The F u n d a m e n t a l C o u n t i n g P r i n c i p l e ) . L e t G a c t o n Q,
a>

Theorem

a n d suppose t h a t O i s one of t h e o r b i t s . L e t a G O, a n d w r i t e H = G t h e s t a b i l i z e r of a . L e t A = { H x | x G G } be t h e set of r i g h t cosets of H i n G. T h e n t h e r e i s a b i j e c t i o n 6 : A -> O such t h a t 9 { H g ) = a-g. I n p a r t i c u l a r , \ 0 \ = \ G : G \.

a

P r o o f . W e observe first t h a t i f H x = H y , t h e n a-x = a-y. T o see w h y this is so, observe t h a t we c a n w r i t e y = hx for some element h s H . T h e n a-y = a - { h x ) = { a - h ) - x = a-x ,
a

where the last equality holds because h G H = G ,

a n d so h stabilizes a .

G i v e n a coset H x G A , the point a-x lies i n O, a n d we know t h a t it is determined by the coset H x , a n d not just b y the p a r t i c u l a r element x . It is therefore permissible to define the function 9 : A O by 9 { H x ) = a-x, and it remains to show t h a t 9 is b o t h injective a n d surjective. The surjectivity is easy, a n d we do t h a t first. If 0 G O, then by the definition of an orbit, we have 0 = a-x for some element x G G. T h e n H x G A satisfies 6 { H x ) = a-x = 0, as required. a-x To prove that 9 is injective, suppose t h a t 9 { H x ) = 9 { H y ) . = a-y, and hence a = a-1 = { a - x ) - x 1 1

W e have

= (a-y)-x" =
a

a-iyx' ).

T h e n yx' fixes a , and so it lies i n G = H . It follows t h a t y G H x , and thus H y = H x . T h i s proves t h a t 6 is injective, as required. It is easy to check t h a t the bijection 9 of the previous theorem a c t u a l l y defines a " p e r m u t a t i o n i s o m o r p h i s m " between the a c t i o n of G on A and the action of G on the orbit O. Formally, this means t h a t 0 ( X - g ) = 9 ( X ) - g for a l l "points" X i n A and group elements g G G. M o r e informally, this says t h a t the actions of G on A and on O are "essentially the same". Since every action can be thought of as composed of the actions on the i n d i v i d u a l orbits, and each of these actions is p e r m u t a t i o n isomorphic to the rightm u l t i p l i c a t i o n action of G on the right cosets of some subgroup, we see t h a t these actions on cosets are t r u l y fundamental: every group action can be viewed as being composed of actions on right cosets of various subgroups. We close this section w i t h two familiar a n d useful applications of the fundamental counting principle.

1. S y l o w

Theory

1.5. C o r o l l a r y . L e t x G, w h e r e G i s a finite g r o u p , a n d l e t K c o n j u g a c y class of G c o n t a i n i n g x . T h e n \ K \ = \ G : C ( x ) \ .

G

be t h e

P r o o f . T h e class of x is the orbit of x under the conjugation action of G on itself, and the stabilizer of x i n this action is the centralizer C { x ) . T h u s \ K \ = \ G : C { x ) \ , as required.
G G

1.6. C o r o l l a r y . L e t H C G be a s u b g r o u p , w h e r e G i s finite. T h e n t h e t o t a l n u m b e r of d i s t i n c t c o n j u g a t e s of H i n G, c o u n t i n g H i t s e l f , i s \ G : N { H ) \ .
G

P r o o f . T h e conjugates of H form a n orbit under the conjugation action of G on the set of subsets of G. T h e normalizer N { H ) is the stabilizer of H in this action, a n d thus the orbit size is \ G : N ( # ) | , as wanted.
G G

Problems

1A

1 A . 1 . L e t H be a subgroup of p r i m e index p i n the finite group G, and suppose that no prime smaller t h a n p divides \ G \ . P r o v e that H < G. 1 A . 2 . G i v e n subgroups H , K C G and an element g G G, the set H g K = { h g k \ h e H , k G K } is called an ( H , i ^ - d o u b l e coset. I n the case where H and K are finite, show that \ H g K \ = \ H \ \ K \ / \ K n H \ .
9

H i n t . Observe t h a t H g K is a u n i o n of right cosets of H , and t h a t these cosets form an orbit under the action of K . N o t e . If we take g = 1 i n this problem, the result is the familiar formula \HK\ = \H\\K\/\HnK\. 1 A.3. Suppose that G is finite a n d t h a t H , K C G are subgroups. (a) Show that \ H : H n K \ < \ G : K \ , w i t h equality if and o n l y if H K = G. (b) If \ G : H \ a n d \ G : K \ are coprime, show t h a t H K = G. N o t e . Proofs of these useful facts appear i n the appendix, but we suggest t h a t readers t r y to find their o w n arguments. A l s o , recall t h a t the product H K of subgroups H and K is not always a subgroup. In fact, H K is a subgroup if a n d o n l y i f H K = K H . (This too is proved i n the appendix.) If H K = K H , we say that H and K are p e r m u t a b l e . 1 A . 4 . Suppose t h a t G = H K , where H and K are subgroups. Show that also G = H K for a l l elements x,y G G. Deduce t h a t if G = H H for a subgroup H a n d a n element x G G, then H = G.
x y X

Problems 1 A

1 A . 5 . A n action of a group G on a set ft is transitive i f ft consists of a single orbit. Equivalently, G is transitive on ft i f for every choice of points a,/? G ft, there exists an element g G G such t h a t a - g = 0. N o w assume t h a t a group G acts t r a n s i t i v e l y on each of two sets ft a n d A . Prove t h a t the n a t u r a l induced action of G on the cartesian p r o d u c t ft x A is transitive if and only i f G G p = G for some choice of a G ft a n d 0 G A .
a

H i n t . Show that if G G p = G for some a G ft and 0 G A , then i n fact, this holds for a l l a G ft and 0 G A .
a

1 A . 6 . L e t G act on ft, where b o t h G a n d ft are finite. F o r each element g G G, w r i t e ( g ) = \ { a e n \ a - g = a } \ . T h e nonnegative-integer-valued function is called the p e r m u t a t i o n character associated w i t h the action. Show that
x X

J ]x(0)

J2

\Ga\=n\G\,

where n is the number of orbits of G on ft. N o t e . T h u s the number of orbits is

geG

w h i c h is the average value of x over the group. A l t h o u g h this orbit-counting formula is often a t t r i b u t e d to W . B u r n s i d e , it s h o u l d (according to P . N e u mann) more p r o p e r l y be credited to C a u c h y a n d Frobenius. 1 A . 7 . L e t G be a finite group, and suppose that H < G is a proper subgroup. Show that the number of elements of G t h a t do not lie i n any conjugate of H is at least \ H \ . H i n t . L e t be the p e r m u t a t i o n character associated w i t h the cation action of G on the right cosets of H . T h e n x ( # ) = sum runs over g e G . Show that x W > 2 | t f | , where runs over h e H . Use this information to get a n estimate o n elements of G where vanishes.
X x

right-multipli\ G \ , where the here, the s u m the number of

1 A . 8 . L e t G be a finite group, let n > 0 be a n integer, and let C be the additive group of the integers m o d u l o n . L e t ft be the set of n-tuples [x ,x ,...,x )
l 2 n

of elements of G such t h a t x x
x

= l.

(a) Show t h a t C acts on ft according to the formula ( x i , x , , x ) - k = ( x i

2 n + k

, x

2 +

k , ,

) ,

where k G C and the subscripts are interpreted m o d u l o n .

1. S y l o w

Theory

(b) N o w suppose t h a t n = p is a prime number t h a t divides \ G \ . Show that divides the number of C - o r b i t s of size 1 on ft, and deduce t h a t the number of elements of order p i n G is congruent to - 1 mod p .
V

N o t e . I n p a r t i c u l a r , if a p r i m e p divides | G | , then G has at least one element of order p . T h i s is a theorem of Cauchy, and the proof i n this p r o b l e m is due to J . H . M c K a y . C a u c h y ' s theorem can also be derived as a corollary of Sylow's theorem. A l t e r n a t i v e l y , a proof of Sylow's theorem different from W i e l a n d t ' s can be based on C a u c h y ' s theorem. (See P r o b l e m 1B.4.) 1 A . 9 . Suppose \ G \ = p m , where p > m and p is prime. Show t h a t G has a unique subgroup of order p . 1A.10. Let H C G . (a) Show t h a t \ N ( H ) : H \ is equal to the number of right cosets of H in G t h a t are invariant under right m u l t i p l i c a t i o n by H .
G

(b) Suppose t h a t \ H \ is a power of the prime p and t h a t \ G : H \ is divisible by p . Show t h a t \ N { H ) : H \ is divisible b y p .

G

IB F i x a prime number p . A finite group whose order is a power of p is called a pgroup. It is often convenient, however, to use this nomenclature somewhat carelessly, and to refer to a group as a "p-group" even if there is no p a r t i c u l a r prime p under consideration. F o r example, i n proving some theorem, one might say: it suffices to check that the result holds for p-groups. W h a t is meant here, of course, is that it suffices to show t h a t the theorem holds for all p-groups for a l l primes p . We m e n t i o n that, a l t h o u g h i n this book a p-group is required to be finite, it is also possible to define infinite p-groups. T h e more general definition is t h a t a (not necessarily finite) group G is a p-group if every element of G has finite p-power order. O f course, i f G is finite, then by Lagrange's theorem, every element of G has order d i v i d i n g \ G \ , and so i f \ G \ is a power of p, it follows t h a t the order of every element is a power of p, a n d hence G is a p-group according to the more general definition. Conversely, if G is finite and has the property t h a t the order of every element is a power of p, then clearly, G can have no element of order q for any prime q different from p. It follows by C a u c h y ' s theorem ( P r o b l e m 1A.8) t h a t no prime q ^ p can divide | G | , and thus \ G \ must be a power of p, and this shows that the two definitions of "p-group" are equivalent for finite groups. A g a i n , fix a prime p. A subgroup S o f a finite group G is said to be a Sylow p-subgroup of G i f \S\ is a power of p and the index \ G : S\ is

IB

not divisible by p. A n alternative formulation of this definition relies on the observation t h a t every positive integer can be (uniquely) factored as a power of the given p r i m e p times some integer not divisible by p . In p a r t i c u l a r , if we w r i t e \ G \ = p m , where a > 0 and p does not divide m > 1, then a subgroup S of G is a Sylow p-subgroup of G precisely w h e n \S\ = p . I n other words, a Sylow p-subgroup of G is a p-subgroup S whose order is as large as is p e r m i t t e d by Lagrange's theorem, w h i c h requires that \S\ must divide | G | . W e m e n t i o n two t r i v i a l cases: i f | G | is not divisible by p, then the i d e n t i t y subgroup is a Sylow p-subgroup of G, a n d if G is a p-group, then G is a Sylow p-subgroup of itself. T h e Sylow existence theorem asserts t h a t Sylow subgroups a l w a y s exist.
a a

1.7. T h e o r e m (Sylow E ) . L e t G be a T h e n G has a S y l o w p - s u b g r o u p .

finite

g r o u p , a n d l e t p be a p r i m e .

The Sylow E - t h e o r e m can be viewed as a p a r t i a l converse of Lagrange's theorem. Lagrange asserts that i f i f is a subgroup of G a n d \ H \ = k, t h e n k divides \ G \ . T h e converse, w h i c h i n general is false, w o u l d say that i f k is a positive integer t h a t divides then G has a subgroup of order k. (The smallest example of the failure of this assertion is to take G to be the alternating group A of order 12; this group has no subgroup of order 6.) But i f A; is a power of a prime, we shall see t h a t G a c t u a l l y does have a subgroup of order k. If k is the largest power of p t h a t divides \ G \ , the desired subgroup of order k is a Sylow p-subgroup; for smaller powers of p, we w i l l prove t h a t a Sylow p-subgroup of G necessarily has a subgroup of order k.
A

We are ready now to begin work t o w a r d the proof of the Sylow E theorem. W e start w i t h a purely a r i t h m e t i c fact about b i n o m i a l coefficients. 1.8. L e m m a . L e t p be a p r i m e n u m b e r , a n d l e t a > 0 a n d m > 1 be i n t e g e r s . Then = m mod p

P r o o f . C o n s i d e r the p o l y n o m i a l (1 + X ) . Since p is prime, it is easy to see t h a t the b i n o m i a l coefficients ( ) are divisible by p for 1 < i < p - 1, and thus we can write (1 + X ) = 1 + X m o d p. ( T h e assertion t h a t these p o l y n o m i a l s are congruent m o d u l o p means t h a t the coefficients of corresponding powers of X are congruent m o d u l o p.) A p p l y i n g this fact a second t i m e , we see t h a t { 1 + X ) = (1+X ) = 1+X m o d p. C o n t i n u i n g like this, we deduce t h a t (1 + X ) = 1+ X m o d p, and thus
p p p p 2 P P p 2 p a p a

( l + Xf

am

= (1 + X

mod p .

10

1. S y l o w

Theory

Since these p o l y n o m i a l s are congruent, the coefficients of corresponding terms are congruent m o d u l o p, and the result follows b y considering the coefficient of X ? on each side.
a

P r o o f of the Sylow E - t h e o r e m ( W i e l a n d t ) . W r i t e \ G \ = p m , where a > 0 a n d p does not divide m. L e t f t be the set of a l l subsets of G having c a r d i n a l i t y p , and observe t h a t G acts by right m u l t i p l i c a t i o n on f t . Because of this action, f t is p a r t i t i o n e d into orbits, and consequently, \fl\ is the s u m of the orbit sizes. B u t
a

and so |i2| is not divisible by p , and it follows that there is some orbit O such t h a t \ 0 \ is not divisible by p . Now let X G O, a n d let H = G be the stabilizer of X i n G. B y the fundamental counting principle, \ 0 \ = \ G \ / \ H \ , a n d since p does not divide \ 0 \ and p divides \ G \ , we conclude that p must d i v i d e \ H \ , and i n particular p < \ H \ .
x a a a

Since H stabilizes X under right m u l t i p l i c a t i o n , we see t h a t if x G X , t h e n x H C X , and thus \ H \ = \ x H \ < \X\ = p , where the final equality holds since X G f t . W e now have \ H \ = p , and since H is a subgroup, it is a Sylow subgroup of G, as wanted.
a a

In P r o b l e m 1 A . 8 , we sketched a proof of C a u c h y ' s theorem. W e c a n now give another proof, using the Sylow E-theorem. 1.9. C o r o l l a r y ( C a u c h y ) . L e t G be a finite g r o u p , a n d suppose of o r d e r p . that p is a

p r i m e d i v i s o r of \ G \ . T h e n G has a n element

P r o o f . L e t S be a Sylow p-subgroup of G, and note that since \S\ is the m a x i m u m power of p that divides | G | , we have \S\ > 1. Choose a nonidentity element x of S, and observe t h a t the order o { x ) divides \S\ by Lagrange's theorem, a n d thus 1 < o ( x ) is a power of p . I n particular, we can w r i t e o { x ) = pm for some integer m > 1, and we see t h a t o { x ) = p , as wanted.
m

We introduce the n o t a t i o n S y l ( G ) to denote the set of a l l Sylow psubgroups of G. T h e assertion of the Sylow E-theorem, therefore, is t h a t the set Sylp(G) is nonempty for a l l finite groups G a n d a l l primes p . T h e intersection n S y l ( G ) of a l l Sylow p-subgroups of a group G is denoted O ( G ) , a n d as we shall see, this is a subgroup that plays an i m p o r t a n t role in finite group theory.
p p p

IB

11

Perhaps this is a good place to digress to review some basic facts about characteristic subgroups. (Some of this m a t e r i a l also appears i n the appendix.) F i r s t , we recall the definition: a subgroup K C G is characteristic in G i f every a u t o m o r p h i s m of G maps K onto itself. It is often difficult to find a l l automorphisms of a given group, and so the definition of "characteristic" can be h a r d to apply directly, but nevertheless, in m a n y cases, it easy to establish t h a t certain subgroups are characteristic. For example, the center Z ( G ) , the derived (or commutator) subgroup G', and the intersection of all Sylow p-subgroups O ( G ) are characteristic i n G . M o r e generally, any subgroup that can be described unambiguously as " t h e something" is characteristic. It is essential t h a t the description using the definite article be unambiguous, however. G i v e n a subgroup H Q G , for example, we cannot conclude t h a t the normalizer N ( H ) or the center Z ( H ) is characteristic i n G . A l t h o u g h these subgroups are described using "the", the descriptions are not unambiguous because they depend on the choice of H . W e can say, however, that Z ( G ' ) is characteristic i n G because it is t h e center of t h e derived subgroup; it does not depend on any unspecified subgroups.
p G

A good way to see w h y "the something" subgroups must be characteristic is to imagine two groups G i and G , w i t h an i s o m o r p h i s m 9 : G i -> G . Since isomorphisms preserve "group theoretic" properties, it should be clear that 9 maps the center Z ( G i ) onto Z ( G ) , and indeed 9 maps each unambiguously defined subgroup of G i onto the corresponding subgroup of G . N o w specialize to the case where G i and G h a p p e n to be the same group G , so 9 is an a u t o m o r p h i s m of G . Since i n the general case, we know that 6 ( Z ( G J ) = Z ( G ) , we see t h a t when G i = G = G , we have 0(Z (G)) = Z ( G ) , and similarly, if we consider any "the something" subgroup i n place of the center.
2 2 2 2 2 2 2

Of course, characteristic subgroups are a u t o m a t i c a l l y n o r m a l . T h i s is because the definition of n o r m a l i t y requires o n l y t h a t the subgroup be m a p p e d onto itself by i n n e r automorphisms while characteristic subgroups are m a p p e d onto themselves by a l l automorphisms. W e have seen that some characteristic subgroups are easily recognized, and it follows that these subgroups are obviously and a u t o m a t i c a l l y n o r m a l . F o r example, the subgroup O (G) is n o r m a l i n G for all primes p .
p

The fact that characteristic subgroups are n o r m a l remains true i n an even more general context. T h e following, w h i c h we presume is already k n o w n to most readers of this book, is extremely useful. (This result also appears i n the appendix.) 1.10. L e m m a . L e t K C N C G, w h e r e G i s a g r o u p , N i s a n o r m a l s u b g r o u p of G a n d K i s a c h a r a c t e r i s t i c s u b g r o u p of N . T h e n K < G .

12

I . Sylow

Theory

P r o o f . L e t g G. T h e n conjugation by g maps N onto itself, and it follows t h a t the restriction of this conjugation m a p to N is an a u t o m o r p h i s m of N . ( B u t note t h a t it is not necessarily an inner a u t o m o r p h i s m of N . ) Since K is characteristic i n N , it is m a p p e d onto itself by this a u t o m o r p h i s m of N , and thus K = K , and it follows t h a t K < G. U
9

Problems

IB
p

1 B . 1 . L e t S e S y l ( G ) , where G is a finite group. (a) L e t P C G be a p-subgroup. only i f P C S.

p

Show that PS

is a subgroup if and

(b) If S < G, show that S y l ( G ) = { S } , and deduce t h a t S is characteristic i n G. N o t e . O f course, it w o u l d be "cheating" to do problems i n this section using theory t h a t we have not yet developed. I n particular, y o u should avoid using the Sylow C-theorem, w h i c h asserts that every two Sylow p-subgroups of G are conjugate i n G. 1B.2. Show t h a t O ( G ) is the unique largest n o r m a l p-subgroup of G. (This means t h a t it is a n o r m a l p-subgroup of G that contains every other n o r m a l p-subgroup of G.)
p

1 B . 3 . L e t S S y l p ( G ) , and write N = N ( 5 ) . Show t h a t N =

G

N (N).
G

1B.4. L e t P C G be a p-subgroup such that \ G : P \ is divisible by p. U s i n g C a u c h y ' s theorem, but w i t h o u t appealing to Sylow's theorem, show that there exists a subgroup Q of G containing P , and such t h a t \ Q : P \ = p . Deduce that a m a x i m a l p-subgroup of G (which obviously must exist) must be a Sylow p-subgroup of G. H i n t . Use P r o b l e m 1A.10 a n d consider the group N ( P ) / P .
G

N o t e . Once we k n o w C a u c h y ' s theorem, this p r o b l e m yields an alternative proof of the Sylow E - t h e o r e m . O f course, to avoid circularity, we appeal to P r o b l e m 1 A . 8 for C a u c h y ' s theorem, and not to C o r o l l a r y 1.9. 1B.5. L e t 7 T be any set of p r i m e numbers. W e say that a finite group H is a vr -group i f every p r i m e divisor of \ H \ lies i n TT . A l s o , a vr-subgroup H C G is a H a l l 7r-subgroup of G i f no p r i m e d i v i d i n g the index \ G : H \ lies i n TT. (So i f TT = {p}, a H a l l vr-subgroup is exactly a Sylow p-subgroup.) N o w let 9 : G - K be a surjective h o m o m o r p h i s m of finite groups. (a) If H is a H a l l vr-subgroup of G, prove that 9 ( H ) is a H a l l vr-subgroup of K .

Problems I B

13

(b) Show that every Sylow p-subgroup of K has the form 9 ( H ) , H is some Sylow p-subgroup of G. (c) Show that | S y l ( G ) | > | S y l ( X ) l for every p r i m e p.
p p

where

N o t e . If the set vr contains more t h a n one p r i m e number, then a H a l l vrsubgroup can fail to exist. B u t a theorem of P . H a l l , after w h o m these subgroups are named, asserts that i n the case where G is solvable, H a l l TTsubgroups always do exist. (See C h a p t e r 3, Section C . ) W e m e n t i o n also that P a r t (b) of this p r o b l e m would not r e m a i n true if "Sylow p-subgroup" were replaced by " H a l l 7r-subgroup". 1B.6. L e t G be a finite group, and let K C G be a subgroup. Suppose that H C G is a H a l l 7r-subgroup, where TT is some set of primes. Show t h a t i f H K is a subgroup, t h e n H n K is a H a l l 7r-subgroup of K . N o t e . In p a r t i c u l a r , K has a H a l l 7r-subgroup if either H or K is n o r m a l i n G since i n that case, H K is guaranteed to be a subgroup. 1 B . 7 . L e t G be a finite group, a n d let vr be any set of primes. (a) Show that G has a (necessarily unique) n o r m a l 7r-subgroup N such t h a t N D M whenever M < G is a 7r-subgroup. (b) Show that the subgroup N of P a r t (a) is contained i n every H a l l 7r-subgroup of G. (c) A s s u m i n g t h a t G has a H a l l 7r-subgroup, show t h a t N is exactly the intersection of a l l of the H a l l vr-subgroups of G. N o t e . T h e subgroup N of this p r o b l e m is denoted O ( G ) . Because of the uniqueness i n (b), it follows t h a t this subgroup is characteristic i n G . F i n a l l y , we note t h a t i f p is a prime number, then, of course, 0 ( G ) - O ( G ) .
n { p } p

1 B . 8 . L e t G be a finite group, and let vr be any set of primes. (a) Show t h a t G has a (necessarily unique) n o r m a l subgroup N such t h a t G / N is a vr-group and M D N whenever M < G a n d G / M is a 7r-group. (b) Show that the subgroup N of P a r t (a) is generated by the set of a l l elements of G that have order not divisible by any p r i m e i n TT. N o t e . T h e characteristic subgroup N of this p r o b l e m is denoted O ( G ) . A l s o , we recall t h a t the subgroup generated by a subset of G is the (unique) smallest subgroup that contains t h a t set.
f f

14

I. Sylow

Theory

1C We are now ready to study i n greater detail the nonempty set S y l ( G ) of Sylow p-subgroups of a finite group G .
p

1.11. T h e o r e m . L e t P be a n a r b i t r a r y p - s u b g r o u p of a suppose t h a t S S y l ( G ) . T h e n PCS f o r some element

9 p

finite g r o u p G, a n d g e G .

P r o o f . L e t ft = {Sx \ x e G } , the set of right cosets of S i n G , and note that |ft| = \G:S\ is not divisible by p since -S is a Sylow p-subgroup of G . We k n o w t h a t G acts by right m u l t i p l i c a t i o n on ft, and thus P acts too, and ft is p a r t i t i o n e d into P - o r b i t s . A l s o , since |ft| is not divisible by p, there must exist some P - o r b i t O such that \ 0 \ is not divisible by p . B y the fundamental counting principle, \ 0 \ is the index i n P of some subgroup. It follows t h a t \ 0 \ divides | P | , w h i c h is a power of p. T h e n \ 0 \ is b o t h a power of p and not divisible by p, and so the only possibility is t h a t \ 0 \ = 1. R e c a l l i n g that a l l members of ft are right cosets of S i n G , we can suppose t h a t the unique member of O is the coset Sg. Since Sg is alone i n a P - o r b i t , it follows that it is fixed under the a c t i o n of P , and thus Sgu = Sg for a l l elements u e P . T h e n g u Sg, and hence g-^Sg = S . T h u s PCS , as required.
9 9 u

If S is a Sylow p-subgroup of G , and g e G is arbitrary, then the conjugate S is a subgroup h a v i n g the same order as S. Since the only requirement on a subgroup that is needed to qualify it for membership i n the set S y l ( G ) is that it have the correct order, and since S S y l ( G ) and \S \ = \S\, it follows t h a t S also lies i n S y l ( G ) . In fact every member of S y l ( G ) arises this way: as a conjugate of S. T h i s is the essential content of the Sylow conjugacy theorem. P u t t i n g it another way: the conjugation a c t i o n of G on S y l ( G ) is transitive.
9 p 9 p 9 p p p

1.12. T h e o r e m (Sylow C ) . If S a n d T S y l o w p - s u b g r o u p s of a G, then T = S

9

finite

group

f o r some

element

g e G .

P r o o f . A p p l y i n g T h e o r e m 1.11 w i t h T i n place of P , we conclude that T C S for some element g e G . B u t since b o t h S and T are Sylow psubgroups, we have \T\ = \S\ = \S \, and so the containment of the previous sentence must a c t u a l l y be an equality.
9 9

The Sylow C-theorem yields an alternative proof of P r o b l e m I B . 1(b), w h i c h asserts that i f a group G has a n o r m a l Sylow p-subgroup S, then S is the only Sylow p-subgroup of G . Indeed, by the Sylow C-theorem, if T e S y l ( G ) , then we c a n w r i t e T = S = S, where the second equality is a consequence of the n o r m a l i t y of S.
9 p

1C

15

A frequently used a p p l i c a t i o n of the Sylow C - t h e o r e m is the so-called " F r a t t i n i argument", w h i c h we are about to present. Perhaps the reason that this result is generally referred to as a n "argument" rather t h a n as a "lemma" or "theorem" is that variations on its proof are used nearly as often as its statement. 1.13. L e m m a ( F r a t t i n i A r g u m e n t ) . L e t N < G w h e r e N i s pose t h a t P G Sylp(iV). T h e n G = N { P ) N .
G 9 9 9

finite,

and

sup-

P r o o f . L e t g G G, and note that P C N = N , and thus P is a subgroup of N h a v i n g the same order as the Sylow p-subgroup P . It follows that P G Sylp(iV), and so by the Sylow C-theorem a p p l i e d i n N , we deduce that ( p g y = p , for some element n J V . Since P = P , we have g n G N ( P ) , and so g G N ( P ) n " C N ( P ) N . B u t g G was arbitrary, and we deduce that G = N ( P ) A T , as required.
9 9 n G 1 G G G

B y definition, a Sylow p-subgroup of a finite group G is a p-subgroup that has the largest possible order consistent w i t h Lagrange's theorem. B y the Sylow E-theorem, we can make a stronger statement: a subgroup whose order is m a x i m a l among the orders of a l l p-subgroups of G is a Sylow psubgroup. A n even stronger assertion of this type is t h a t every m a x i m a l psubgroup of G is a Sylow p-subgroup. Here, " m a x i m a l " is to be interpreted in the sense of containment: a subgroup H of G is m a x i m a l w i t h some property i f there is no subgroup K > H t h a t has the property. T h e t r u t h of this assertion is the essential content of the Sylow "development" theorem. 1.14. T h e o r e m (Sylow D ) . L e t P be a p - s u b g r o u p of a finite g r o u p G. P i s c o n t a i n e d i n some
p

Then

S y l o w p - s u b g r o u p of G.
9

P r o o f . L e t S G S y l ( G ) . T h e n by T h e o r e m 1.11, we k n o w that PCS for some element g G G. A l s o , since \S \ = \S\, we know t h a t S is a Sylow p-subgroup of G. U
9 9

G i v e n a finite group G, we consider next the question of how m a n y Sylow p-subgroups G has. T o facilitate this discussion, we introduce the (not quite standard) n o t a t i o n n { G ) = | S y l ( G ) | . (Occasionally, when the group we are considering is clear from the context, we w i l l s i m p l y write n instead of n {G).)
p p p p

F i r s t , by the Sylow C-theorem, we k n o w t h a t S y l ( G ) is a single orbit under the conjugation action of G. T h e following is then an immediate consequence.
p

1.15. C o r o l l a r y . L e t S G S y l ( G ) , w h e r e G i s a finite g r o u p .
p

Thenn (G)
p

|G:N (5)|.
G

16

1. S y l o w

Theory

P r o o f . Since n ( G ) = | S y l ( G ) | is the t o t a l number of conjugates of S i n G, the result follows by C o r o l l a r y 1.6.

p p

In particular, it follows that n ( G ) divides \ G \ , but we can say a bit more. If 5 S y l ( G ) , then of course, S C N ( S ) since S is a subgroup, and thus | G : S\ = \ G : N ( 5 ) | | N ( 5 ) : S\. A l s o , n { G ) = \ G : N ( 5 ) | , and hence n ( G ) divides \ G : S\. I n other words, if we write \ G \ = p m , where p does not divide m , we see that n { G ) divides m. (We mention that the integer m is often referred to as the p '-part of \ G \ . )
p p G G G p G a p p

T h e information that n { G ) divides the p'-part of \ G \ becomes even more useful w h e n it is combined w i t h the fact (probably k n o w n to most readers) t h a t n ( G ) = 1 m o d p for a l l groups G. In fact, there is a useful stronger congruence constraint, w h i c h may not be quite so well k n o w n . Before we present our theorem, we m e n t i o n that i f S,T S y l ( G ) , then |5| = | T | , and thus \S : SnT\ = \S\/\SnT\ = | T | / | 5 n r | = \ T : S n T \ . T h e statement of the following result, therefore, is not really as asymmetric as it may appear.
p p p

1.16. T h e o r e m . Suppose t h a t G i s a finite g r o u p such t h a t n { G ) > 1 , a n d choose d i s t i n c t S y l o w p - s u b g r o u p s S a n d T of G such t h a t t h e o r d e r \S D T\ i s as l a r g e as p o s s i b l e . T h e n n { G ) = 1 m o d \S : S n T | .

p p

1.17. C o r o l l a r y . If G i s a mod p.
p

finite

group and p is a prime, then n ( G ) = 1

p

P r o o f . If n ( G ) = 1, there is n o t h i n g to prove. Otherwise, T h e o r e m 1.16 applies, a n d there exist distinct members S,T e S y l ( G ) such that n ( G ) = 1 m o d \S :SnT\, and thus it suffices to show t h a t \S :SnT\ is divisible by p. B u t \S : S n T\ = \ T : S n T\ is certainly a power of p, and it exceeds 1 since otherwise S = S n T = T , w h i c h is not the case because S and T are distinct.
p p

In order to see how T h e o r e m 1.16 can be used, consider a group G of order 21,952 = 2 - 7 . W e know that n must divide 2 = 64, and it must be congruent to 1 m o d u l o 7. W e see, therefore, that n must be one of 1, 8 or 64. Suppose that G does not have a n o r m a l Sylow 7-subgroup, so that n > 1. Since neither 8 nor 64 is congruent to 1 m o d u l o 7 = 49, we see by T h e o r e m 1.16 that there exist distinct Sylow 7-subgroups S and T of G such that \S : SnT\ = 7.
6 3 6 7 7 2 7

L e t ' s pursue this a b i t further. W r i t e D = S ( I T i n the above situation, and note t h a t since |5 : D \ = 7 is the smallest prime divisor of \S\ = 7 , it follows by P r o b l e m 1 A . 1 , t h a t D < S. S i m i l a r reasoning shows t h a t also D < T , a n d hence S and T are b o t h contained i n N = N ( L > ) . N o w S and T are distinct Sylow 7-subgroups of N , and it follows t h a t n ( N ) > 1, a n d hence n ( N ) > 8 by C o r o l l a r y 1.17. Since n ( N ) is a power of 2 that
3 G 7 7 7

1C

17

divides | AT |, we deduce t h a t 2 have \ G : N \ < 8.

divides \ N \ .

Since also 7

divides \ N \ , we

We can use what we have established to show that a group G of order 21,952 cannot be simple. Indeed, if n ( G ) = 1, then G has a n o r m a l subgroup of order 7 , and so is not simple. Otherwise, our subgroup N has index at most 8, and we see t h a t | G | does not d i v i d e \ G : N \ \ . B y the n \ theorem ( C o r o l l a r y 1.3), therefore, G cannot be simple i f N < G. F i n a l l y , \{ N = G then D < G and G is not simple i n this case too.
7 3

In the last case, where D < G, we see t h a t D is contained i n a l l Sylow 7-subgroups of G, and thus D is the intersection of every two distinct Sylow 7-subgroups of G. I n most situations, however, T h e o r e m 1.16 can be used to prove only the existence of some pair of distinct Sylow subgroups w i t h a "large" intersection; it does not usually follow t h a t every such pair has a large intersection. To prove T h e o r e m 1.16, we need the following. 1.18. L e m m a . L e t P S y l ( G ) , w h e r e G i s a that Q is a p-subgroup o/N (P). Then Q C P .
p G

finite

group, and

suppose

P r o o f . W e a p p l y Sylow theory i n the group N = N ( P ) . Clearly, P is a Sylow p-subgroup of N , and since P < N , we deduce that P is the only Sylow p-subgroup of N . B y the Sylow D - t h e o r e m , however, the p-subgroup Q of N must be contained i n some Sylow p-subgroup. T h e only possibility is Q C P , as required.
G

A n alternative m e t h o d of proof for L e m m a 1.18 is to observe that since Q C ^ N ( P ) , it follows that Q P = P Q . T h e n Q P is a subgroup, and it is easy to see that it is a p-subgroup t h a t contains the Sylow p-subgroup P . It follows t h a t P = Q P 5 Q, as wanted.
G

P r o o f of T h e o r e m 1.16. L e t S act on the set S y l ( G ) by conjugation. One orbit is the set { S } , of size 1, a n d so i f we can show t h a t a l l other orbits have size divisible by \S : S D T\, it w i l l follow t h a t n ( G ) = | S y l ( G ) | = 1 mod \S : S D T\, as wanted. L e t O be an a r b i t r a r y S - o r b i t i n S y l ( G ) other t h a n { S } and let P G O, so that P ^ S. B y the fundamental counting p r i n ciple, \ 0 \ = \S : Q\, where Q is the stabilizer of P i n S under conjugation. T h e n Q C N ( P ) , and so Q C P by L e m m a 1.18. B u t also Q C 5 , and thus | Q | < | 5 n P | < | 5 n T | ; where the latter inequality is a consequence of the fact that |5 D T\ is as large as possible a m o n g intersections of two distinct Sylow p-subgroups of G . It follows that \ 0 \ = \S : Q\ > \S : S D T\. B u t since the integers \ 0 \ and \S : S n T\ are powers of p and \ 0 \ > \S : S n T\, we conclude that \ 0 \ is a m u l t i p l e of \S : S n T\. T h i s completes the proof.
p p p p G

18

1. S y l o w

Theory

Problems

IC
p G

1 C . 1 . L e t P e S y l ( G ) , and suppose that N ( P ) C H C G , where H is a subgroup. P r o v e that H = N { H ) .

G

N o t e . T h i s generalizes P r o b l e m 1 B . 3 . 1 C . 2 . L e t H C G , where G is a finite group. (a) If P S y l ( P " ) , prove that P = H n 5 for some member 5 G Syl (G).
p p

(b) Show that n ( P f ) < n ( G ) for a l l primes p.

p p

1 C . 3 . L e t G be a finite group, and let X be the subset of G consisting of all elements whose order is a power of p , where p is some fixed prime. (a) Show that X = ( J S y l ( G ) .
p

1 C . 4 . L e t \ G \ = 120 = 2 -3-5. Show that G has a subgroup of index 3 or a subgroup of index 5 (or b o t h ) . H i n t . A n a l y z e separately the four possibilities for n ( G ) .
2

1 C . 5 . L e t P e S y l p ( G ) , where G = A the alternating group on p + 1 symbols. Show that | N ( P ) | = p(p - l ) / 2 .

p + U G

H i n t . C o u n t the elements of order p i n G . 1 C . 6 . L e t G = PTif, where P" and K are subgroups, and fix a prime p. (a) Show that there exists P S y l ( G ) such that P n H S y l ( P ) and P H f i S y l ( X ) .
p p p

(b) If P is as i n (a), show that P - ( P n H ) ( P n AT). H i n t . F o r (a), first choose Q S y l ( G ) and g G such that Q n H S y l ( P ) and Q n f i e S y l ( X ) . W r i t e 0 = Ziifc, w i t h h e H and k <E K .
p 9 p p

1 C . 7 . L e t G be a finite group i n w h i c h every m a x i m a l subgroup has prime index, a n d let p be the largest p r i m e divisor of | G | . Show that a Sylow p-subgroup of G is n o r m a l . H i n t . Otherwise, let M be a m a x i m a l subgroup of G containing N ( P ) , where P e S y l ( G ) . C o m p a r e n { M ) and n ( G ) .
G p p p

ID

19

1 C . 8 . L e t P be a Sylow p-subgroup of G. Show t h a t for every nonnegative integer a, the numbers of subgroups of order p i n P a n d i n G are congruent modulo p .
a

N o t e . If p = | P | , then the number of subgroups of order p i n P is clearly 1, a n d it follows that the number of such subgroups i n G is congruent to 1 m o d u l o p . T h i s provides a somewhat different proof that n ( G ) = 1 m o d p . It is true i n general that i f p < \P\, then the number of subgroups of order p i n P is congruent to 1 m o d u l o p , a n d thus it follows t h a t i f p divides the order of an a r b i t r a r y finite group G, t h e n the n u m b e r of subgroups of order p i n G is congruent to 1 m o d p .
p a a a a

ID W e now digress from our study of Sylow theory i n order to review some basic facts about p-groups a n d nilpotent groups. A l s o , we discuss the F i t t i n g subgroup, a n d i n the problems at the end of the section, we present some results about the F r a t t i n i subgroup. A l t h o u g h p-groups are not at a l l t y p i c a l of finite groups i n general, they play a prominent role i n group theory, a n d they are ubiquitous i n the s t u d y of finite groups. T h i s u b i q u i t y is, of course, a consequence of the Sylow theorems, a n d perhaps t h a t justifies our digression. W e should m e n t i o n t h a t although their structure is a t y p i c a l when compared w i t h finite groups i n general, p-groups are, nevertheless, extremely abundant i n comparison w i t h non-p-groups. T h e r e are, for example, 2,328 i s o m o r p h i s m types of groups of order 128 = 2 ; the number of types of order 256 = 2 is 56,092; for 512 = 2 the number is 10,494,213; a n d there are exactly 49,487,365,422 i s o m o r p h i s m types of groups of order 1,024 = 2 . (These numbers were c o m p u t e d by a remarkable a l g o r i t h m for counting pgroups t h a t was developed by E . O ' B r i e n . )
7 8 9 1 0

T h e r e is an extensive theory of finite p-groups (and also of their infinite cousins, pro-p-groups), and there are several books entirely devoted to t h e m . O u r brief presentation here w i l l be quite superficial; later, we study p-groups a b i t more deeply, but s t i l l , we shall see only a t i n y part of what is k n o w n . Perhaps the most fundamental fact about p-groups is that n o n t r i v i a l finite p-groups have n o n t r i v i a l centers. ( B y our definition, "p-group" means "finite p-group", but we included the redundant adjective i n the previous sentence a n d i n what follows i n order to stress the fact that finiteness is essential here. Infinite p-groups can have t r i v i a l centers, a n d i n fact, they can be simple groups.) In fact, a stronger statement is true.

20

1. S y l o w

Theory

1.19. T h e o r e m . L e t P be a finite p - g r o u p a n d l e t N be a n o n i d e n t i t y n o r m a l s u b g r o u p of P . T h e n N n Z ( P ) > 1. I n p a r t i c u l a r , if P i s n o n t n v i a l , t h e n Z { P ) > 1. P r o o f . Since N < P , we can let P act on N by conjugation, and we observe that N f l Z ( P ) is exactly the set of elements of N that lie i n orbits of size 1. B y the fundamental counting principle, every orbit has p-power size, and so each n o n t r i v i a l orbit ( i . e . , orbit of size exceeding 1) has size divisible by p . Since the set N - ( N n Z ( P ) ) is a u n i o n of such orbits, we see t h a t \ N \ \ N n Z ( P ) | is divisible by p , and thus \ N n Z ( P ) | = \ N \ = 0 m o d p , where the second congruence follows because TV is a n o n t r i v i a l subgroup. Now N n Z ( P ) contains the identity element, and so \ N n Z ( P ) | > 0. It follows t h a t \ N n Z ( P ) | > p > 1, and hence N n Z ( P ) is n o n t r i v i a l , as required. T h e final assertion follows by t a k i n g N = P . U It is now easy to show that (finite, of course) p-groups are nilpotent, and thus we can o b t a i n a d d i t i o n a l information about p-groups by s t u d y i n g general nilpotent groups. B u t first, we review some definitions. A finite collection of n o r m a l subgroups N of a (not necessarily finite) group G is a n o r m a l series for G provided that
t

1 = iVo C N i C C N

= G.

T h i s n o r m a l series is a central series if i n a d d i t i o n , we have N i / N i - i C Z ( G / J V i _ i ) for 1 < i < r . F i n a l l y , a group G is nilpotent if it has a central series. It is w o r t h noting t h a t subgroups and factor groups of nilpotent groups are themselves nilpotent, although we omit the easy proofs of these facts. G i v e n any group G, we can a t t e m p t to construct a central series as follows. ( B u t of course, this attempt is doomed to failure unless G is nilpotent.) We start by defining Z = 1 and Z = Z ( G ) . T h e second center Z is defined to be the unique subgroup such t h a t Z / Z = Z { G / Z ) . (Note that Z exists and is n o r m a l i n G by the correspondence theorem.) W e continue like this, i n d u c t i v e l y defining Z for n > 0 so that Z / Z - i = Z ( G / Z _ i ) . The chain of n o r m a l subgroups
0 x 2 2 1 1 2 n n n

1 = Z

C Z iC Z

constructed this way is called the u p p e r central series of G. W e hasten to point out, however, t h a t i n general, the upper central series may not actually be a central series for G because it may happen that Z < G for a l l i. In other words, the upper central series may never reach the whole group G. B u t if Z = G for some integer r , then { Z \ 0 < i < r } is a true central series, and G is nilpotent.
{ r x

ID

21

Conversely, i f G is nilpotent, the upper central series of G really is a central series. F o r finite groups G, this is especially easy to prove. 1.20. L e m m a . L e t G be (1) G i s n i l p o t e n t . (2) E v e r y n o n t r i v i a l h o m o m o r p h i c i m a g e of G has a n o n t r i v i a l (3) G a p p e a r s as a member of i t s u p p e r c e n t r a l s e r i e s . center. finite. Then the following a r e equivalent.

P r o o f . W e have already remarked t h a t h o m o m o r p h i c images of nilpotent groups are nilpotent. A l s o , since the first n o n t r i v i a l t e r m of a central series for a nilpotent group is contained i n the center of the group, it follows t h a t n o n t r i v i a l nilpotent groups have n o n t r i v i a l centers. T h i s shows t h a t (1) implies (2). A s s u m i n g (2) now, it follows t h a t i f Z < G, where is a t e r m i n the upper central series for G, then Z / Z i = Z { G / Z i ) is n o n t r i v i a l , and thus Zi < Z . Since G is finite and the proper terms of the upper central series are s t r i c t l y increasing, we see t h a t not every t e r m can be proper, and this establishes (3).
{ l + 1 i + 1

F i n a l l y , (3) guarantees t h a t the upper central series for G is actually a central series, and thus G is nilpotent, p r o v i n g (1). If P is a finite p-group, then of course, every h o m o m o r p h i c image of P is also a finite p-group, and thus every n o n t r i v i a l h o m o m o r p h i c image of P has a n o n t r i v i a l center. It follows by L e m m a 1.20, therefore, t h a t finite p-groups are nilpotent. I n fact, we shall see i n T h e o r e m 1.26 t h a t m u c h more is true: a finite group G is nilpotent i f and o n l y i f every Sylow subgroup of G is normal. N e x t , we show t h a t the terms of the upper central series of a nilpotent group contain the corresponding terms of an a r b i t r a r y central series, and this explains w h y the upper central series is called "upper". It also provides an alternative proof of the i m p l i c a t i o n (1) => (3) of L e m m a 1.20, w i t h o u t the assumption t h a t G is finite. 1.21. central Theorem. series l = N a n d as u s u a l , l e t 1 = Z be t h e u p p e r c e n t r a l series p a r t i c u l a r , Z = G.
r 0 Q

L e t G be a ( n o t n e c e s s a r i l y C N C . . - C N

finite) = G,

nilpotent group with

C Z

C Z

C
t

f o r G.

Then N

C Z

for 0 < %< r, and i n

22

1. S y l o w

Theory

P r o o f . W e prove that N i C Z i by i n d u c t i o n on i. Since Z = 1 = A^o, we can suppose that i > 0, and by the inductive hypothesis, we can assume t h a t N i - i C Z i - i . F o r n o t a t i o n a l simplicity, write N = and Z = Z - , and observe t h a t since N C Z , there is a n a t u r a l surjective h o m o m o r p h i s m 0 : G / N G / Z , defined by 0 ( W ) = Z<? for elements g e G. (This m a p is well defined since Z g is the unique coset of Z t h a t contains N g , and so 0 { N g ) depends o n l y on the coset N g and not on the element g . )
0 % X 5

Since 9 is surjective, it carries central elements of G / N to central elements of G / Z , and since N / N is central i n G/N, it is m a p p e d by 0 into Z ( G / Z ) = Z i / Z . If x N therefore, it follows that Zx = 0 { N x ) 6 Z / Z , and thus x E Z as required.
t i } x h

If G is an a r b i t r a r y nilpotent group, then G is a t e r m of its upper central series, w h i c h , therefore is a true central series. W e have G = Z for some integer r > 0, and the smallest integer r for w h i c h this happens is called the nilpotence class of G . T h u s n o n t r i v i a l abelian groups have nilpotence class 1, and the groups of nilpotence class 2 are exactly the nonabelian groups G such t h a t G / Z ( G ) is abelian.
r

Since the upper central series of the nilpotent group G is really a central series, we see t h a t i f G has nilpotence class r , then it has a central series of length r . (In other words, the series has r containments and r + 1 terms.) B y T h e o r e m 1.21, we see that G cannot have a central series of length smaller t h a n r, and thus the nilpotence class of a nilpotent group is exactly the length of its shortest possible central series. I n some sense, the nilpotence class c a n be viewed as a measure of how far from being abelian a nilpotent group is. O n e of the most useful facts about nilpotent groups, and thus also about finite p-groups, is t h a t "normalizers grow". 1.22. T h e o r e m . L e t H < G , w h e r e G i s a ( n o t n e c e s s a r i l y group. Then N ( H ) > H .
G

finite)

nilpotent

Before we proceed w i t h the (not very difficult) proof of T h e o r e m 1.22, we digress to discuss the "bar convention", w h i c h provides a handy n o t a t i o n for dealing w i t h factor groups. Suppose that N < G, where G is an a r b i t r a r y group. W e write G to denote the factor group G/N, and we t h i n k of the overbar as the name of the canonical h o m o m o r p h i s m from G onto G . If g e G is any element, therefore, we write g to denote the image of g i n G , and so we see t h a t g is s i m p l y another name for the coset N g . A l s o , i f H C G is any subgroup, t h e n H is the image of H i n G . B y the correspondence theorem, the h o m o m o r p h i s m "overbar" defines a bijection from the set of those subgroups of G t h a t contain A^ onto the set of all subgroups of G . E v e r y subgroup of G , therefore, has the form H for

ID

23

some subgroup H Q G , and although there_are usually m a n y subgroups of G whose image i n G is the given subgroup H , e x a c t l y o n e o f t h e m contains N . If H Q G is arbitrary, we see t h a t H N = H N = H since overbar is a h o m o m o r p h i s m and N is its kernel. _It follows t h a t H N is the unique subgroup c o n t a i n i n g N whose image i n G is H . I n p a r t i c u l a r , since indices of corresponding subgroups are equal, we have \ G : H \ = \ G : A^P"! for a l l subgroups H Q G . The correspondence theorem also yields information_about normality. < K i f and o n l y _if H < K . I n p a r t i c u l a r ,
G G

If AT C H Q K C G, then i f N C H , N
G

t h e n since 7f C N ( / f ) , we see t h a t # < N ( P ) and we have I n fact, equality holds here. T o see this, observe t h a t U Q G. T h e n H < U, and so H < U and
G

( H )

C N q ( H ) .

since N g ( f f ) is a subgroup of G, it can be w r i t t e n in_the form U for some (unique) subgroup U w i t h N Q

G

U C N ( f f ) . T h i s yields N ^ T ? ) = t7 C N ( 7 f ) , as c l a i m e d . W e w i l l use the bar convention i n the following proof.

P r o o f of T h e o r e m 1.22. Since G is nilpotent, it has (by definition) a cent r a l series { N i | 0 < i < r } , and we have NQ = 1 C H and N = G H . It follows t h a t there is some subscript k w i t h 0 < k < r such t h a t N C H but N % H . W e w i l l show t h a t i n fact, A ^ C N ( f f ) , and it w i l l follow t h a t N ( f f ) > H , as required.
r k k + 1 f c + 1 G G

W r i t e G = G/N and use the bar convention. Since the subgroups N i form a central series, we have
k

NkVi C Z ( G )C N ^ H )= N G W ) , where the equality holds because N C H . N o w because N C N ( f f ) , we can remove the overbars to o b t a i n A T i C N ( H ) . T h e proof is now complete.
k k G f c + G

We r e t u r n now to p-groups, w i t h another a p p l i c a t i o n of T h e o r e m 1.19. 1.23. L e m m a . L e t P be a n o r m a l s u b g r o u p s of P . Then L C M a n d \ L : JV| = p . finite p - g r o u p a n d suppose that N t h e r e exists a s u b g r o u p L < P such < M are that N C

P r o o f . W r i t e J P - P / N and note that M is n o n t r i v i a l and n o r m a l i n P . N o w Z ( P ) i l l is n o n t r i v i a l by T h e o r e m 1.19, and so this subgroup contains an element of order p . (Choose any n o n i d e n t i t y element and take an appropriate power.) Because our element is central and of order p , it generates a n o r m a l subgroup of order p , and we can write I to denote this subgroup, where N C L . N o w I C M , and thus AT C L C M , as wanted. A l s o , as L < P , we see t h a t L < P . F i n a l l y , \ L : N \ = | Z | = p , as required.

24

1. S y l o w

Theory

1.24. C o r o l l a r y . L e t P be a p - g r o u p of o r d e r p . T h e n f o r every w i t h 0 < b < a , t h e r e i s a s u b g r o u p L < P such t h a t \ L \ = p .

b

integer b

P r o o f . T h e assertion is t r i v i a l i f b = 0, and so we can assume t h a t b > 0 and we work by i n d u c t i o n on b. B y the inductive hypothesis, there exists a subgroup N < P such t h a t |7V| = p ~ and we can a p p l y T h e o r e m 1.22 ( w i t h M = P ) to produce a subgroup L < P w i t h \ L : N \ = p . T h e n | L | = p and the proof is complete.
b l b

R e c a l l now that the Sylow E-theorem can be viewed as a p a r t i a l converse to Lagrange's theorem. It asserts t h a t for certain divisors k of an integer n , every group of order n has a subgroup of order k. (The divisors to w h i c h we refer, of course, are prime powers k such t h a t n / k is not divisible by the relevant prime.) We can now enlarge the set of divisors for w h i c h we know t h a t the converse of Lagrange's theorem holds. 1.25. C o r o l l a r y . L e t G be a f i n i t e g r o u p , a n d suppose that p divides \G\, w h e r e p i s p r i m e a n d b > 0 i s a n i n t e g e r . T h e n G has a s u b g r o u p of o r d e r p .
b b

P r o o f . L e t P be a Sylow p-subgroup of G and write \ P \ = p .

Since p

divides \ G \ , we see t h a t b < a , and the result follows b y C o r o l l a r y 1.23.

In fact, i n the s i t u a t i o n of C o r o l l a r y 1.25, the number of subgroups of G h a v i n g order p is congruent to 1 m o d u l o p . ( B y P r o b l e m 1C.8 and the note following it, it suffices to prove this i n the case where G is a p-group, and while this is not especially difficult, we have decided not to present a proof here.) W e m e n t i o n also t h a t it does not seem to be k n o w n whether or not there are any integers n other t h a n powers of primes such t h a t every group of order divisible by n has a subgroup of order n .
b

Sylow theory is also related to the theory of nilpotent groups i n another way: a finite group is nilpotent i f and only if a l l of its Sylow subgroups are n o r m a l . I n fact, we can say more. 1.26. T h e o r e m . L e t G be a finite g r o u p . T h e n t h e f o l l o w i n g a r e (1) G i s n i l p o t e n t . (2) N
G

equivalent.

( H ) > H f o r every

p r o p e r s u b g r o u p H < G.

(3) E v e r y m a x i m a l s u b g r o u p of G i s n o r m a l . (4) E v e r y S y l o w s u b g r o u p of G i s n o r m a l . (5) G i s t h e ( i n t e r n a l ) d i r e c t p r o d u c t of i t s n o n t r i v i a l S y l o w subgroups.

ID

25

N o t e t h a t i n statement (3), a " m a x i m a l subgroup" is m a x i m a l among p r o p e r subgroups. B u t i n most other situations, the w o r d " m a x i m a l " does not i m p l y proper. If a group G happens to be nilpotent, for example, then the whole group is a m a x i m a l nilpotent subgroup of G. To help w i t h the proof that (4) implies (5), we establish the following. 1.27. L e m m a . L e t X be a c o l l e c t i o n of finite n o r m a l s u b g r o u p s of a g r o u p G, a n d assume t h a t t h e o r d e r s of t h e members of X a r e p a i r w i s e c o p r i m e . Then the product H = \ \ X of t h e members of X i s d i r e c t . A l s o , \ H \ = T\ xex \X\.

P r o o f . C e r t a i n l y \ H \ < U\ \A l s o , by Lagrange's theorem, \X\ divides \ H \ , for every member X of X , and since the orders of the members of X are pairwise coprime, it follows t h a t f j \X\ divides \ H \ . W e conclude t h a t \H\ = ni^l, as wanted. Now to see t h a t T \ X direct, it suffices to show t h a t

X n Y[{Y e x |Y + X }= l for every member X EX. T h i s follows since by the previous paragraph, the

order of n Y for Y ^ X is equal to n \Y\, and this is coprime to | X | .

P r o o f of T h e o r e m 1.26. W e saw t h a t (1) implies (2) i n T h e o r e m 1.22. T h a t (2) implies (3) is clear, since i f M < G is a m a x i m a l subgroup, then N (M) > M , and so we must have N ( M ) = G.
G G

Now assume (3), and let P E S y \ ( G ) for some p r i m e p . If N ( P ) is proper i n G, it is contained i n some m a x i m a l subgroup M , and we have M < G. Since P S y l ( M ) , it follows by L e m m a 1.13, the F r a t t i n i argument, t h a t G = N ( P ) M C M , and this is a c o n t r a d i c t i o n . T h u s P < G, and this proves (4).
p G p G

T h a t (4) implies (5) is immediate from L e m m a 1.27. N o w assume (5). It is clear t h a t (4) holds, and it follows t h a t (4) also holds for every homomorphic image of G. (See P r o b l e m 1B.5.) Since we k n o w t h a t (4) implies (5), we see t h a t every h o m o m o r p h i c image of G is a direct p r o d u c t of p-groups for various primes p . T h e center of a direct p r o d u c t , however, is the d i rect product of the centers of the factors, and since n o n t r i v i a l p-groups have n o n t r i v i a l centers, it follows t h a t every nonidentity h o m o m o r p h i c image of G has a n o n t r i v i a l center. W e conclude by L e m m a 1.20 t h a t G is nilpotent, thereby establishing (1). R e c a l l t h a t O ( G ) is the unique largest n o r m a l p-subgroup of G, by w h i c h we m e a n t h a t it contains every n o r m a l p-subgroup. W e define the F i t t i n g s u b g r o u p of G, denoted F ( G ) , to be the p r o d u c t of the subgroups
p

26

1. S y l o w

Theory

O { G ) as p runs over the prime divisors of G. O f course, F ( G ) is characterp

istic i n G, and i n particular, it is n o r m a l . 1.28. C o r o l l a r y . L e t G be a finite g r o u p . T h e n F ( G ) i s a n o r m a l n i l p o t e n t s u b g r o u p of G. I t c o n t a i n s every n o r m a l n i l p o t e n t s u b g r o u p of G, a n d so i t i s t h e u n i q u e l a r g e s t such subgroup. P r o o f . B y L e m m a 1.27, we k n o w that | F ( G ) | is the product of the orders of the subgroups O ( G ) as p runs over the prime divisors of G . T h e n O ( G ) S y l ( F ( G ) ) , and thus F ( G ) has a n o r m a l Sylow subgroup for each prime. It follows by T h e o r e m 1.26 t h a t F ( G ) is nilpotent.
p p p

N o w let N < G be nilpotent. If P e S y l ( i V ) , then P < N by T h e o r e m 1.26, and thus P is characteristic i n N and hence is n o r m a l i n G . It follows t h a t P C O ( G ) C F ( G ) . Since N is the product of its Sylow subgroups and each of these is contained i n F ( G ) , it follows t h a t N C F ( G ) , and the proof is complete.
p p

1.29.

Corollary. L e t K

a n d L be n i l p o t e n t n o r m a l s u b g r o u p s

of a

finite

g r o u p G. T h e n K L i s n i l p o t e n t . P r o o f . W e have K C F ( G ) and L C F ( G ) , and thus K L C F ( G ) . Since F ( G ) is nilpotent, it follows t h a t K L is nilpotent.

Problems

ID
p G

1 D . 1 . L e t P 6 S y l ( P ) , where H C G , and suppose t h a t N ( P ) C H . Show t h a t p does not divide | G : H \ . N o t e . I n these problems, we are, as usual, dealing w i t h finite groups. 1D.2. F i x a p r i m e p , and suppose t h a t a subgroup H C G has the property t h a t C ( x ) C H for every element x E H h a v i n g order p . Show that p cannot divide b o t h | P | and | G : P | .
G

1D.3. L e t H C G have the property that H n P < /< EG P. (a) Show t h a t N

G

= 1 for all elements

( K ) C P for a l l subgroups # w i t h

K i C i i .

(b) Show t h a t H is a H a l l subgroup of G . (Recall t h a t this means t h a t \ H \ and | G : P | are coprime.) N o t e . A subgroup H < G that satisfies the hypothesis of this p r o b l e m is usually referred to as a F r o b e n i u s complement i n G . In C h a p t e r 6, we give a different definition of a "Frobenius complement", w h i c h , i n fact, is equivalent to this one. It is easy to see (as we w i l l show) t h a t a Frobenius complement i n the sense of C h a p t e r 6 satisfies the hypothesis of this p r o b l e m .

Problems I D

27

W h a t is m u c h more difficult is the theorem of Frobenius, w h i c h asserts t h a t if H is a Frobenius complement i n G i n the sense denned here, then it is, in fact, a Frobenius complement i n the sense of C h a p t e r 6. Unfortunately, all k n o w n proofs of Frobenius' theorem require character theory, and so we cannot present a proof i n this book. 1D.4. L e t G = N H , where I < N < G and N n H 1 . Show t h a t H is a Frobenius complement i n G (as denned above) i f and o n l y i f C ( h ) = 1 for all nonidentity elements h H .
N

H i n t . If x and x

lie i n H , where n N , observe t h a t x ~ x

N .

1D.5. L e t H < G, and suppose t h a t N ( P ) C H for a l l p-subgroups P C H , for all primes p . Show that H is a Frobenius complement i n G .
G

H i n t . Observe t h a t the hypothesis is satisfied by H n H for 6 G. If this intersection is n o n t r i v i a l , consider a n o n t r i v i a l Sylow subgroup Q o i H n H and show t h a t Q and Q " conjugate i n H .
9 5 9 1 a r e

1D.6. Show t h a t a subgroup of a nilpotent group is m a x i m a l if and o n l y i f it has p r i m e index.

1D.7. F o r any finite group G, the F r a t t i n i s u b g r o u p $ ( G ) is the intersection of all m a x i m a l subgroups. Show t h a t $ ( G ) is exactly the set of "useless" elements of G , by w h i c h we mean the elements g e G such t h a t if ( X U { g } } = G for some subset X of G , t h e n ( X ) = G.

1D.8. A finite p-group is elementary a b e l i a n i f it is abelian and every nonidentity element has order p . If G is nilpotent, show t h a t G / $ ( G ) is abelian, and t h a t if G is a p-group, then G / $ ( G ) is elementary abelian. N o t e . It is not h a r d to see t h a t i f P is a p-group, then $(P) is the unique n o r m a l subgroup of P m i n i m a l w i t h the property t h a t the factor group is elementary abelian. 1D.9. If P is a noncyclic p-group, show t h a t \ P : $ ( P ) | > p , and deduce that a group of order p
2 2

must be either cyclic or elementary abelian.

I D . 1 0 . L e t A be m a x i m a l among the abelian n o r m a l subgroups of a p-group P . Show t h a t A = C { A ) , and deduce t h a t \ P : A \ divides ( \ A \ - 1)!.
P

Hint. Let C = C { A ) .
P

If C > A , a p p l y L e m m a 1.23.

28

1. S y l o w

Theory

1 D . 1 1 . L e t n be the m a x i m u m of the orders of the abelian subgroups of a finite group G . Show t h a t | G | divides n \ . H i n t . Show t h a t for each prime p , the order of a Sylow p-subgroup of G divides n \ . N o t e . There exist infinite groups i n w h i c h the abelian subgroups b o u n d e d order, so finiteness is essential here. I D . 1 2 . L e t p be a p r i m e d i v i d i n g the order of a group G. Show t h a t the number of elements of order p i n G is congruent to - 1 m o d u l o p . I D . 1 3 . If Z C Z ( G ) and G / Z is nilpotent, show t h a t G is nilpotent. I D . 1 4 . Show t h a t the F r a t t i n i subgroup $ ( G ) of a finite group G is nilpotent. H i n t . A p p l y the F r a t t i n i argument. T h e proof here is somewhat similar to the proof t h a t (3) implies (4) i n T h e o r e m 1.26. N o t e . T h i s p r o b l e m shows that $ ( G ) C F ( G ) . I D . 1 5 . Suppose t h a t $ ( G ) C N < G and that N / $ ( G ) is nilpotent. Show t h a t N is nilpotent. I n particular, if G / $ ( G ) is nilpotent, then G is nilpotent. N o t e . T h i s generalizes the previous p r o b l e m , w h i c h follows b y setting N = $(G). N o t e t h a t this p r o b l e m proves t h a t F ( G / $ ( G ) ) = F ( G ) / $ ( G ) . I D . 1 6 . L e t N < G, where G is finite. Show t h a t $(JV) C $ ( G ) . H i n t . If some m a x i m a l subgroup M of G fails to contain $ ( j V ) , then have

$ ( W ) A f = G , and it follows that N = < P ( N ) ( N n M ) . 1D.17. L e t N < G, where AT is nilpotent and G/AT' G is nilpotent. is nilpotent. P r o v e t h a t

H i n t . T h e derived subgroup N ' is contained i n $(JV) by P r o b l e m 1D.8. N o t e . If we weakened the assumption that G/N' is nilpotent and assumed instead that G / N is nilpotent, it w o u l d not follow t h a t G is necessarily nilpotent. 1D.18. Show that F ( G / Z ( G ) ) = F ( G ) / Z ( G ) for a l l finite groups G . I D . 1 9 . L e t F = F ( G ) , where G is an a r b i t r a r y finite group, and let C = C { F ) . Show t h a t G / ( G D F ) has no n o n t r i v i a l abelian n o r m a l subgroup.
G

H i n t . Observe t h a t F ( G ) < G .

IE

29

IE F r o m the earliest days of group theory, researchers have been i n t r i g u e d by the question: what are the finite simple groups? O f course, the abelian simple groups are exactly the groups of p r i m e order, and (up to isomorphism) there is just one group of order p for each p r i m e p : the cyclic group of t h a t order. B u t n o n a b e l i a n simple groups are c o m p a r a t i v e l y rare. T h e r e are, for example, o n l y five numbers less t h a n 1,000 t h a t occur as orders of n o n a b e l i a n simple groups, and up to i s o m o r p h i s m , there is just one simple group of each of these orders. (These numbers are 60, 168, 360, 504 and 660.) T h e r e do exist numbers, however, such t h a t there are two nonisomorphic simple groups of t h a t order. (The smallest such number is 8!/2 = 20,160.) B u t no number is the order of three nonisomorphic simple groups. Perhaps their r a r i t y is one reason t h a t n o n a b e l i a n finite simple groups have inspired such intense interest over the years. It seems quite n a t u r a l to collect rare objects and to attempt to acquire a complete collection. B u t a more " p r a c t i c a l " e x p l a n a t i o n is t h a t a knowledge of all finite simple groups and their properties w o u l d be a major step i n understanding a l l finite groups. The reason for this is that, i n some sense, all finite groups are b u i l t from simple groups. To be more precise, suppose t h a t G is any n o n t r i v i a l finite group. B y finiteness, G has at least one m a x i m a l n o r m a l subgroup N . (We mean, of course, a m a x i m a l p r o p e r n o r m a l subgroup, but as is customary i n this context, we have not made.the word "proper" explicit.) T h e n by the correspondence theorem, the group G / N is simple. ( T h i s is because the n o r m a l subgroups of G / N are i n n a t u r a l correspondence w i t h the n o r m a l subgroups of G that c o n t a i n N , and there are just two of these: N and G.) N o w i f N is n o n t r i v i a l , we can repeat the process by choosing a m a x i m a l n o r m a l subgroup M of N . (In general, of course, M w i l l not be n o r m a l i n G.) Because G is finite, we see t h a t i f we continue like this, repeatedly choosing a maxi m a l n o r m a l subgroup of the previously selected group, we must eventually reach the identity subgroup. If we number our subgroups from the b o t t o m up, we see that we have constructed (or more accurately "chosen") a c h a i n of subgroups N such that 1 = jV < N
0 x

< < N

= G

such t h a t each of the factors A ^ - i is simple for 1 < i < r . I n this situation, the subgroups iV are said to form a c o m p o s i t i o n series for G, and the simple groups N / N - x are the corresponding c o m p o s i t i o n factors. T h e J o r d a n - H o l d e r theorem asserts t h a t despite the arbitrariness of the c o n s t r u c t i o n of the c o m p o s i t i o n series, the set of c o m p o s i t i o n factors (including multiplicities) is uniquely determined up to i s o m o r p h i s m . T h e
?

30

1. S y l o w

Theory

c o m p o s i t i o n factors of G are the simple groups from w h i c h we might say t h a t G is constructed. (We mention t h a t the finite groups for w h i c h all c o m p o s i t i o n factors are cyclic of prime order are exactly the "solvable" finite groups, w h i c h we s t u d y i n more detail i n C h a p t e r 3.) The p r o b l e m of finding all nonabehan finite simple groups can be approached from two directions: construct as m a n y simple groups as y o u can, and prove t h a t every finite nonabelian simple group appears i n your list. In p a r t i c u l a r , as part of the second of these programs, it is useful to prove "nonsimplicity" theorems that show t h a t groups that look different from the k n o w n simple groups cannot, i n fact, be simple. A major result of this type from early i n the 20th century is due to W . B u r n s i d e , who showed t h a t the order of a nonabelian simple group must have at least three different prime divisors. ( T h i s is B u r n s i d e ' s classic pY-theorem.) B u r n s i d e also observed t h a t all of the then k n o w n nonabelian simple groups had even order. He conjectured that this holds i n general: t h a t all odd-order simple groups are cyclic of prime order. B u r n s i d e ' s conjecture, w h i c h can be paraphrased as the assertion t h a t every group of o d d order is solvable, was eventually proved by W . Feit and J . G . T h o m p s o n i n the early 1960s. (The celebrated F e i t - T h o m p s o n paper, at about 250 pages, may have been the longest published proof of a single theorem at the time.) Since around 1960, there has been d r a m a t i c progress w i t h b o t h aspects of the simple-group-classification problem, and it appears t h a t now, i n the early years of the 21st century, the classification of finite simple groups is complete. So what are the nonabehan finite simple groups? A highly abbreviated description is this. E v e r y finite nonabelian group is either: (1) O n e of the alternating groups A
n

f o r n > 5,

(2) A member of one of a number of infinite families parameterized by prime-powers q and (usually) by integers n > 2, or (3) One of 26 other "sporadic" simple groups t h a t do not fit into types (1) or (2). Of the simple groups i n parameterized families, the easiest to describe are the projective special linear groups P S L ( n , q ) , where q is a primepower and n > 2. These are constructed as follows. L e t F be the field of order q and construct the general linear group G L ( n , q ) consisting of all invertible n x n matrices over F . T h e special linear group S L ( n , q ) is the n o r m a l subgroup of G L { n , q ) consisting of those matrices w i t h determinant 1. It is not h a r d to see that the center Z = Z ( S L ( n , q ) ) consists exactly of the scalar matrices of determinant 1, and by definition, P S L ( n , q ) is the factor group S L ( n , q ) / Z . It turns out that P S L ( n , q ) is simple except

IE

31

w h e n n = 2 and q is 2 or 3. (We say more about the groups P S L ( n , q ) i n C h a p t e r 7, and we prove their s i m p l i c i t y i n C h a p t e r 8, where we also prove t h a t the alternating groups A are simple for n > 5.)
n

In fact, a l l of the simple groups w i t h order less t h a n 1,000 are of the form P S L ( 2 , q ) , where q is one of 5, 7, 8, 9 or 11, a n d the corresponding group orders are 60, 168, 504, 360 and 660. T h e unique (up to isomorphism) simple group P S L { 2 , 5) of order 60 has two other realizations: it is isomorphic to P S L ( 2 , 4 ) and also to the a l t e r n a t i n g group A . T h e simple groups P S L ( 2 , 7) of order 168 and P S L ( 2 , 9) of order 360 also have m u l t i p l e realizations: the first of these is isomorphic to P S L ( 3 , 2) and the second is isomorphic to A .
5 6

The smallest of the 26 sporadic simple groups is the s m a l l M a t h i e u group, denoted M , of order 7,920; the largest is the Fischer-Griess "monster" of order
n

2 -3 -5 -7 -ll -17-19-23-29-31-41.47-59-71 = 808,017,424,794,512,875,886,459,904,961,710,757,005,754,368,000,000,000. For the remainder of this section, we discuss n o n s i m p l i c i t y theorems of the form: "if the order of G is . . . , then G cannot be s i m p l e " . ( O f course, b o t h B u r n s i d e ' s p Y - t h e o r e m and the F e i t - T h o m p s o n odd-order theorem are of this type.) A very m u c h more elementary result of this form is i m m e d i a t e from the fact t h a t a n o n t r i v i a l p-group always has a n o n t r i v i a l center. If | G | is divisible by only one prime, it follows t h a t G cannot be simple unless Z (G) = G , and i n this case, G is abelian, a n d so it must be cyclic of p r i m e order. B u r n s i d e ' s pV-theorem asserts t h a t the order of a simple group cannot have e x a c t l y two p r i m e divisors. H i s beautiful (and short) proof is not elementary since it uses character theory (which we do not discuss i n this b o o k ) . It took about 50 years before a purely group-theoretic (and harder) proof of B u r n s i d e ' s theorem was found b y G o l d s c h m i d t , M a t s u y a m a and Bender, using powerful techniques of T h o m p s o n , G l a u b e r m a n and others. These techniques were developed for other purposes, and eventually they led to the full classification of finite simple groups, but as a test of their power, it seemed reasonable to see if they w o u l d y i e l d a direct proof of B u r n s i d e ' s theorem. Indeed they d i d , and one of the goals of this b o o k is to develop enough group theory so that we can present a somewhat simplified version of the G o l d s c h m i d t - M a t s u y a m a - B e n d e r proof of B u r n s i d e ' s theorem. ( T h i s proof appears i n C h a p t e r 7.) For now, however, we use Sylow theory (and a few tricks) to prove some m u c h more elementary n o n s i m p l i c i t y theorems.

46

20

32

1. S y l o w

Theory

1.30. T h e o r e m . L e t \ G \ = p q , w h e r e q < p a r e p r i m e s . T h e n G has a n o r m a l S y l o w p - s u b g r o u p . A l s o , G i s c y c l i c unless q divides p - 1 . P r o o f . Since n = 1 m o d p , we see t h a t i f n > 1, we have n > p > q. T h i s is not possible, however, since n must divide q. W e conclude t h a t n = 1, as wanted. A l s o , since every element of G of order p generates a subgroup of order p , a n d there is only one of these, there are e x a c t l y p - 1 elements of order p i n G.
p p p p p

Now if n , > 1, then p = n = 1 m o d q, and thus q divides p - 1 . Otherwise n = 1, and we see t h a t there are exactly q - 1 elements of order q i n G. Together w i t h the identity, we have now accounted for p + q - 1 elements of G. B u t p + q - 1 < 2 p < p q = \ G \ , and so G must have some element g t h a t does not have order either 1, g or p. Since o { g ) divides \ G \ = p q , the o n l y possibility is t h a t o ( g ) = p q , and thus G = {pq) is cyclic.
q q

1.31.

Theorem. L e t \G\ = p q , w h e r e p a n d q a r e p r i m e s .

T h e n G has

either a n o r m a l Sylow p-subgroup o ra n o r m a l Sylow

q-subgroup.

P r o o f . W e assume t h a t n > 1 a n d n , > 1, and we note t h a t p ^ q. Since n = 1 m o d p , it follows t h a t n > p and s i m i l a r l y n > q.
p p p q

Now n must divide q, and hence n = q and we have n > q = n > p . B u t n divides p , and we see t h a t the only possibility is t h a t n = p , w h i c h means that G has p subgroups of order q.
p p q p 2 2 q q 2

If Q i ^ Q are subgroups of order q, then since q is prime, it follows via Lagrange's theorem t h a t Q n Q = 1. T h e p subgroups of order q in G, therefore, have no nonidentity elements i n c o m m o n , and it follows t h a t G has at least p { q - 1) elements of order q. ( A c t u a l l y , since every element of order q i n G must lie i n some subgroup of order q, it follows t h a t G has e x a c t l y p ( q - l ) elements of order q, b u t we shall not need that fact.) T h e number of elements of G t h a t do not have order q is then at most \ G \ - p { q - 1) = p . If P e S y l p ( G ) , then, of course, no element of P has order q, and since | P | = p , it follows that P is exactly equal to the set of elements of G t h a t do not have order q. I n p a r t i c u l a r , P is uniquely determined, a n d this is a c o n t r a d i c t i o n since we are assuming t h a t n > 1.
2 2 l 2 2 2 2 2 2 p

B y the previous two theorems, we see t h a t if p ^ q are primes, then a group of order p q or of order p q must either have a n o r m a l Sylow p subgroup or a n o r m a l Sylow g-subgroup. T h i s almost works for groups of order p q too, but there is an exception. T h i s s i t u a t i o n is endemic i n finite group theory: some general fact holds w i t h a s m a l l n u m b e r of exceptions.
2 3

IE

33

( A n o t h e r example of this phenomenon is t h a t a l l automorphisms of the s y m m e t r i c group S are inner for a l l integers n > 1 except n = 6.)
n

1.32. T h e o r e m . L e t \ G \ = p q , w h e r e p a n d q a r e p r i m e s . T h e n G has e i t h e r a n o r m a l S y l o w p - s u b g r o u p o r a n o r m a l S y l o w q - s u b g r o u p , except when \ G \ = 24. P r o o f . A s i n the previous proof, we assume t h a t n > 1 and t h a t n > 1, and we conclude t h a t p ^ q and t h a t n > p and n > q. C o n t i n u i n g to reason as we d i d previously, we have n > q = n > p . Since n divides p , we are left this t i m e w i t h two possibilities: either n = p or n = p .
p q p q 3 q p q 3 2 q q

Suppose t h a t n = p . T h e n the p subgroups of order q c o n t a i n a t o t a l of p { q - l ) elements of order q, and thus G c&ntains at most \ G \ - p { q - l ) = p elements t h a t do not have order q. It follows t h a t i f P e S y l ( G ) , then P is exactly the set of elements of G h a v i n g order different from q, and thus P is unique, c o n t r a d i c t i n g the assumption t h a t n > 1.
q 3 3 3 p p

We conclude, therefore, t h a t n = p , a n d thus p = 1 m o d q. In other words, q divides p - 1 = ( p + l ) ( p - 1). Since q is prime, we see t h a t q must d i v i d e either p + 1 or p - 1, and so i n either case, we have q < p + l . B u t we k n o w from the first paragraph of the proof t h a t p < q, and so the only possibility is t h a t q = p + l . N o w p and q are primes t h a t differ by 1, and so one of t h e m must be even. W e deduce t h a t p = 2 a n d q = 3, and thus \ G \ = 24.
q 2

Of course, the previous proof does not t e l l us t h a t there a c t u a l l y is an exception of order 24; it only permits the possibility of such an exception. B u t a n exception really does exist. C o n s i d e r S , the s y m m e t r i c group on four symbols, w h i c h , of course, has order 4! = 24. B y e x a m i n i n g the possible cycle structures for elements of S , it is easy to count t h a t there are exactly nine elements of order 2, eight elements of order 3 a n d six elements of order 4. ( A s a check, note t h a t 9 + 8 + 6 = 23, so t h a t together w i t h the identity, we have accounted for a l l 24 elements.) B y the Sylow D - t h e o r e m , every element of 2-power order i n S lies i n some Sylow 2-subgroup. Since each Sylow 2subgroup has order 8, and yet there are more t h a n eight elements of 2-power order, it follows t h a t S must have more t h a n one Sylow 2-subgroup. S i m i l a r reasoning shows that there must be more t h a t one Sylow 3-subgroup. (In fact, it is easy to see t h a t n (S ) = 3 and t h a t n (S ) = 4.)
4 4 4 4 2 4 3 4

T h e r e is s t i l l more t h a t we can say about this s i t u a t i o n . W e k n o w t h a t among groups of order p q , only groups of order 24 can fail to have a n o n t r i v i a l n o r m a l Sylow subgroup, and we k n o w t h a t at least one group of order 24 a c t u a l l y does fail to have such a Sylow subgroup. T h e r e m a i n i n g question is whether or not there are any other groups of order 24 that can occur as exceptions. T h e answer is "no".
3

34

1. S y l o w

Theory

1.33. Then

Theorem. G^S .
4

L e t \ G \ = 24, a n d suppose

n ( G ) > 1 a n d n ( G ) > 1.
2 3

P r o o f . Since n exceeds 1, is congruent to 1 m o d u l o 3 and divides 8, the only possibility is t h a t n = 4, and thus \ G : N \ = 4, where N = N ( P ) and P S y l ( G ) . N o w let K = c o r e ( N ) , so that G / K is isomorphic to a subgroup of S b y T h e o r e m 1.1. It suffices to show t h a t K = 1 since that w i l l i m p l y t h a t G is isomorphic to a subgroup of 5 , and this w i l l complete the proof since \ G \ = 24 = \S \.
3 3 G 3 G 4 4 A

R e c a l l t h a t K C i V = N ( P ) , where P G S y l ( G ) . T h e n P is a n o r m a l Sylow 3-subgroup of J T P , and so P is characteristic i n K P . Since P is not n o r m a l i n G, however, we conclude t h a t K P cannot be n o r m a l i n G. B u t K P / K is a Sylow 3-subgroup of G/K, and it follows t h a t n { G / K ) > 1. I n particular, G / K is not a 2-group, and so \ K \ is not divisible by 3.
G 3 3

Since \ G : N \ = 4, we see t h a t |TV| = 6, and hence the only possibilities are \ K \ = 1, as wanted, or \ K \ = 2. A s s u m i n g now that \ K \ = 2, we work to o b t a i n a c o n t r a d i c t i o n . Since \ G / K \ = 12, T h e o r e m 1.31 applies, and we deduce t h a t either n { G / K ) = 1 or n { G / K ) = 1. Since we have seen t h a t n { G / K ) > 1, however, we conclude t h a t G / K has a unique Sylow 2-subgroup S/K. T h e n S/K < G/K, and so S < G. A l s o | 5 | = 8 since \ K \ = 2 a n d \S/K\ = 4, and thus S is a n o r m a l Sylow 2-subgroup of G. T h i s contradicts the hypothesis t h a t n { G ) > 1.
2 3 3 2

We mention t h a t the group S , of order 24, is not simple, and it follows t h a t no group of order p q is simple, where p and q are primes. I n fact, S has a n o r m a l subgroup of order 12 and one of order 4. T h e subgroup of order 12 is the alternating group A (see the discussion below) and the n o r m a l subgroup of order 4 is the so-called K l e i n g r o u p consisting of the identity and the three elements of order 2 i n S t h a t have no fixed points.
4 3 4 4 4

W e can eliminate a quarter of a l l positive integers as possibilities for the order of a simple group. T h e argument relies on the n o t i o n of o d d and even permutations, w i t h w h i c h , we assume the reader is familiar. Briefly, the facts are these. E v e r y p e r m u t a t i o n of a finite set can be w r i t t e n as a p r o d u c t of transpositions (2-cycles). A p e r m u t a t i o n t h a t can be w r i t t e n as a p r o d u c t of an even number of transpositions is even and one t h a t can be w r i t t e n as a product of an o d d number of transpositions is o d d . It is a t r i v i a l i t y t h a t the even permutations i n the s y m m e t r i c group S form a subgroup. (It is called the alternating g r o u p and is denoted A . ) It is also true, but m u c h less t r i v i a l to prove, that no p e r m u t a t i o n can be b o t h even a n d o d d , and since o d d permutations certainly exist for n > 2, we see t h a t A is proper i n S i n this case.
n n n n

IE

35

M o r e generally, suppose t h a t G is an a r b i t r a r y group t h a t acts on some finite set ft, and assume t h a t there is at least one element x G G t h a t induces a n o d d p e r m u t a t i o n on ft. It should be clear t h a t the elements of G t h a t induce even permutations on ft form a subgroup, w h i c h we call H . Since x i H , we see t h a t H < G, and we c l a i m t h a t \ G : H \ = 2. T o see this, it suffices to show t h a t the set G - H is contained i n a single right coset of H . W e show, i n fact, t h a t i f g G G - H , t h e n g G E x - . N o w g induces an odd p e r m u t a t i o n on ft, and hence gx induces an even p e r m u t a t i o n and we have gx G H . T h e n g G H x ~ , as c l a i m e d . Since a subgroup of index 2 is a u t o m a t i c a l l y n o r m a l , we have established the following useful fact.
l 1

I. 34. L e m m a . L e t G a c t o n a ofG acts

finite

set ft a n d suppose

t h a t some

element

" o d d l y " . T h e n G has a n o r m a l s u b g r o u p of i n d e x 2.

1.35. T h e o r e m . Suppose t h a t \G\ = I n , w h e r e n i s odd. n o r m a l s u b g r o u p of i n d e x 2.

T h e n G has a

P r o o f . B y C a u c h y ' s theorem, we can find an element t G G of order 2. I n the regular action of G o n itself by right m u l t i p l i c a t i o n , t has no fixed points, a n d so the p e r m u t a t i o n induced by t consists entirely of 2-cycles, and there are \ G \ / 2 = n of t h e m . Since n is o d d , the p e r m u t a t i o n induced b y t is o d d , a n d the result follows by L e m m a 1.34. We have stated t h a t 60 is the smallest number t h a t occurs as the order of a nonabelian simple group. In fact, using w h a t we have now proved, this is not very h a r d to see. F i r s t , we can eliminate a l l powers of primes since i f G is a simple p-group, then because 1 < Z ( G ) < G, we see t h a t Z ( G ) = G and G is abelian. A l s o , by T h e o r e m 1.30, we can eliminate a l l products of two primes, and thus we can assume t h a t | G | factors as a product of at least three primes, counting multiplicity. Suppose now t h a t \ G \ < 60 and write n = \ G \ . If n is o d d , then, since 3 , 3 - 7 and 3-5 a l l exceed 60, we see t h a t the o n l y possibilities that are products of at least three primes are n = 3 and n = 3 - 5 , and we have seen t h a t neither of these numbers can be the order of a simple group.
4 2 2 3 2

We are left w i t h the cases where n is even. B y T h e o r e m 1.35, we can suppose t h a t n is divisible by 4, and we w r i t e n = 4 m , where m < 15. W e have e l i m i n a t e d the cases where m is p r i m e or is a power of 2, and also the cases where m = 2 p , where p is a n o d d p r i m e . T h e the only s u r v i v i n g possibilities for m are 9 and 12, and so we need to show t h a t there is no simple group of order 36 or of order 48. If G is simple of order 36 = 2 - 3 , we see t h a t | G : P \ = 4, where P G S y l ( G ) , and yet | G : P \ does not d i v i d e 4! = 24. T h i s violates the n!-theorem, C o r o l l a r y 1.3. Similarly, i f | G | = 48 = 2 - 3 , we get a v i o l a t i o n of
3 4 2 2

36

1. S y l o w

Theory

the n!-theorem by t a k i n g P S y l ( G ) , so t h a t | G : P \ = 3. T h i s completes the proof t h a t 60 is the smallest possible order of a nonabelian simple group.
2

To prove t h a t no number between 60 and 168 c a n be the order of a simple group, it is convenient to have two more n o n s i m p l i c i t y results of the type we have been discussing: if G has order p q or order p q r , where p , q and r are primes, then G cannot be simple. (We leave the proofs of these to the exercises at the end of this section.) U s i n g these two facts and the other results t h a t we have established, a l l but about a dozen of the numbers between 60 and 168 can be e l i m i n a t e d , and most of those are easily dealt w i t h by a d h o c methods. F o r example, we need to consider the numbers 84 = 2 -3-7, 140 = 2 -5-7 and 156 = 2 -3-13, a n d these are easily e l i m i n a t e d by the observation t h a t i f p is the largest prime divisor, then n = 1 since no divisor of n exceeding 1 is congruent to 1 m o d u l o p .
2 2 2 2 2 p

The number 72 = 2 - 3 can be e l i m i n a t e d by observing t h a t i f G is simple and \ G \ = 72, then n > 1, and therefore the only possibility is n = 4. T h e n \ G : N \ = 4, where N is the normalizer of a Sylow 3subgroup, a n d this violates the n!-theorem. M o r e interesting is the case where \ G \ = 144 = 2 - 3 . T h i s case can be handled using T h e o r e m 1.16, w i t h an argument analogous to the one we used i n the discussion following C o r o l l a r y 1.17, where we showed t h a t groups of order 2 - 7 are not simple. A n o t h e r number t h a t requires some work is 132 = 2 - 3 - l l . If a group of this order is simple, then n = 12, and hence there are at least 120 elements of order 11, leaving at most 12 other elements. It follows that n = 4, and this yields a c o n t r a d i c t i o n using the n!-theorem.
3 3 4 2 6 3 2 n 3

The number between 60 and 168 t h a t seems to be the most difficult to eliminate is 120 = 2 -3-5. One (rather tricky) approach is this. If G is simple of order 120, it is easy to see t h a t n ( G ) = 6. T h e n G acts (nontrivially) on the six right cosets of the normalizer of a Sylow 5-subgroup, a n d this determines a n o n t r i v i a l h o m o m o r p h i s m from G into the s y m m e t r i c group 5 . Since G is simple, this h o m o m o r p h i s m is injective, and thus there is an isomorphic copy H of G w i t h H C S . B u t G has no n o r m a l subgroup of index 2, and hence by L e m m a 1.34, no element of G can act o d d l y on the six cosets, and thus H a c t u a l l y lies i n the a l t e r n a t i n g group A , and since \ A \ = 360, we see t h a t \ A : H \ = 3. R e c a l l , however, t h a t the alternating groups A are simple w h e n n > 5, and so by the n!-theorem, A cannot have a subgroup of index 3. T h i s c o n t r a d i c t i o n shows t h a t 120 is not the order of a simple group.
3 5 6 6 6 6 6 n 6

A n alternative approach for e l i m i n a t i n g 120 is to quote P r o b l e m 1 C . 4 , w h i c h asserts t h a t if \ G \ = 120, then there exists a subgroup H C G w i t h 1 < \ G : H \ < 5. If G is simple, this yields an i s o m o r p h i s m of G into the s y m m e t r i c group S . Since |5 | = 120 = \ G \ , we conclude that G = S ,
5 5 5

Problems I E

37

a n d this is a contradiction since S alternating group A .

5

has a n o r m a l subgroup of index 2: the

We close this section w i t h one further result w h i c h , t h o u g h somewhat more general t h a n those t h a t we have already proved, is s t i l l o n l y a very special case of B u r n s i d e ' s pV-theorem. 1.36. T h e o r e m . Suppose that \G\ = p q , w h e r e p a n d q are p r i m e s a > 0. T h e n G i s n o t s i m p l e .
p a

and

P r o o f . W e c a n certainly assume t h a t p ^ q a n d t h a t n > 1, a n d thus n = q. N o w choose distinct Sylow p-subgroups S a n d T of G such t h a t \S f l T\ is as large as possible, a n d write D = S n T.
p

If D = 1, then every pair of distinct S y l o w p-subgroups of G intersect t r i v i a l l y , a n d so the Sylow p-subgroups of G account for a t o t a l of q { p - 1) nonidentity elements of G. A l l of these elements, of course, have orders divisible b y p , a n d this leaves r o o m for at most \ G \ - q { p - 1) = q elements w i t h order not divisible by p . It follows t h a t i f Q G S y l , ( G ) , then Q is exactly the set of these elements, a n d i n p a r t i c u l a r Q is unique, a n d so Q < G a n d G is not simple, as required.
a q

We c a n now assume t h a t D > 1 , a n d we let N = N ( D ) . Since D < S and D < T a n d we know t h a t " n o r m a l i z e s grow" i n p-groups, we c a n conclude t h a t N Ci S > D a n d N (~]T > D .
G

N e x t , we show t h a t N is not a p-group. Otherwise, b y the Sylow D theorem, we c a n write N C Re S y l ( G ) . T h e n R n S D N n S > D , a n d so b y the choice of D , we see that the Sylow subgroups R a n d S cannot be distinct. In other words, S = R a n d similarly, T = R. B u t this is a c o n t r a d i c t i o n since S ^ T .
p

It follows t h a t q divides | j V | , a n d so i f we choose Q G S y l , ( i V ) , we have | Q | = q. N o w S is a p-group a n d Q is a g-group, a n d thus S n Q = 1 and we have \SQ\ = \S\\Q\ = p q = \ G \ . T h e n SQ = G, a n d hence i f g G G is arbitrary, we c a n write g = xy, where x G S a n d y G Q. T h e n g xy = sv Z} D V = D , where the second equality holds since x G S a n d the last equality follows because y e Q C N = N ( > ) . W e see, therefore, t h a t D is contained i n every conjugate of S i n G. T h e n D is contained i n every Sylow p-subgroup of G, a n d we have 1 < D C O ( G ) . T h u s O ( G ) is a nonidentity proper n o r m a l subgroup G , a n d this completes the proof t h a t G is not simple.
a S = S G p p

Problems I E 1 E . 1 . L e t | G | = p q , where p < q are primes. P r o v e t h a t n { G ) = 1 unless \ G \ = 36.

q 2 2

38

1. S y l o w

Theory

N o t e . If | G | = 3 6 , then a Sylow 3 subgroup really can fail to be n o r m a l , but i n t h a t case, one c a n show t h a t a Sylow 2-subgroup of G is n o r m a l . 1 E . 2 . L e t \ G \ = p q r , where p < q < r are primes. Show that n ( G ) = 1.
r

H i n t . Otherwise, show by c o u n t i n g elements t h a t a Sylow g-subgroup must be n o r m a l and consider the factor group, of order p r . N o t e . I n general, i f \ G \ is a product of distinct primes, t h e n a Sylow subgroup for the largest p r i m e divisor of \ G \ is n o r m a l . W e prove this theorem of B u r n s i d e w h e n we study transfer theory i n C h a p t e r 5. 1 E . 3 . Show t h a t there is no simple group of order 315 = 3 -5-7. N o t e . T h i s , of course, is also a consequence of the F e i t - T h o m p s o n odd-order theorem, w h i c h asserts t h a t no group of o d d nonprime order can be simple. 1 E . 4 . If \ G \ = 144 = 2 - 3 , show that G is not simple. 1 E . 5 . If \ G \ = 336 = 2 -3-7, show t h a t G is not simple. H i n t . If G is simple, compute n ( G ) and use P r o b l e m 1 C . 5 .
7 4 4 2 2

1 E . 6 . If \ G \ = 180 = 2 - 3 - 5 , show t h a t G is not simple. 1 E . 7 . If \ G \ = 240 = 2 -3-5, show that G is not simple. 1 E . 8 . If \ G \ = 252 = 2 - 3 - 7 , show that G is not simple. IF R e c a l l t h a t O ( G ) is the unique largest n o r m a l p-subgroup of the finite group G, and t h a t it can be found by t a k i n g the intersection of a l l of the Sylow p-subgroups of G. B u t do we really need a l l of them? W h a t is the smallest collection of Sylow p-subgroups of G w i t h the property t h a t their intersection is O ( G ) ? In the case where a Sylow p-subgroup is abelian, there is a pretty answer, w h i c h was found by J . S. B r o d k e y . ( O f course, since the Sylow p-subgroups of G are conjugate by Sylow C-theorem, they are isomorphic, and so if any one of t h e m is abelian, they a l l are.)
p p 2 2 4

1.37. T h e o r e m ( B r o d k e y ) . Suppose t h a t a S y l o w p - s u b g r o u p of a finite g r o u p G i s a b e l i a n . T h e n t h e r e e x i s t S,T e S y l ( G ) such t h a t S n T = O ( G ) .

p p

Of course, every intersection of two Sylow p-subgroups of G contains O ( G ) , so what B r o d k e y ' s theorem really says is that if the Sylow subgroups are abelian, then O ( G ) is the unique m i n i m a l such intersection. I n fact, what is essentially B r o d k e y ' s argument establishes something about m i n i m a l intersections of two Sylow p-subgroups even i f the Sylow subgroups are not abelian.
p p

IF

39

1.38. T h e o r e m . F i x a p r i m e p , a n d l e t G be any finite g r o u p . Choose S,T e Sylp(G) such t h a t D = S n T i s m i n i m a l i n t h e set of i n t e r s e c t i o n s of t w o S y l o w p - s u b g r o u p s o f G . T h e n O ( G ) i s t h e u n i q u e l a r g e s t s u b g r o u p of D t h a t i s n o r m a l i n b o t h S a n d T.

p

If we assume that S and T are abelian i n T h e o r e m 1.38, we see that D < S and D < T , and so i n this case, the theorem guarantees t h a t D = O ( G ) . In other words, B r o d k e y ' s theorem is an i m m e d i a t e corollary of T h e o r e m 1.38.
p

P r o o f of T h e o r e m 1.38. L e t K C D , where K < S and K < T. W e must show t h a t K C O ( G ) , or equivalently, t h a t K C P for a l l Sylow p-subgroups P of G.
p

Let N = N { K ) and note that S C N , and thus S G S y l ( i V ) . N o w let P G Sylp(G) and observe t h a t F i l i V i s a p-subgroup of N . T h e n P n i V is contained i n some Sylow p-subgroup of N , a n d so we can w r i t e P f ] N C S for some element x N . (We are, of course, using the Sylow D - a n d C theorems i n the group N . ) N o w T C N , and thus also T C N and we have
G p x x

p n T Then

= P n N r \ T

c s

n T

= D

D = (D ) ~ Since P ~
X X

D ( Pn ^ f '

= P ~

nT .

and T are Sylow p-subgroups of G , and their intersection is

x - 1

contained i n D , it follows by the m i n i m a l i t y of D t h a t P ~ ' n T = D . In p a r t i c u l a r , we have J f C D C P *

G

, a n d thus i f * C P . B u t K finite

= K since

x i V = N ( J f ) , and thus K C P , as required. 1.39.

p

C o r o l l a r y . L e i P G S y l ( G ) , w/tere G is a
2 p

group, and

assume

that P is abelian. Then |G : O ( G ) \ < \G : P| . P r o o f . B y B r o d k e y ' s theorem, we can choose S , T G S y l ( G ) such t h a t

p

S H T

= O ( G ) . Then
p

| G |

"

\snr\

\O (G)\
p

T h i s yields | G | / | P | > l / | O ( G ) | , and the result follows by m u l t i p l y i n g b o t h sides of this inequality by | G | .

p

We

can recast this as a "good" n o n s i m p l i c i t y theorem. finite and P is abelian, and G is not simple unless

1.40. C o r o l l a r y . L e t P G S y l ( G ) , w h e r e G i s
p

assume |G| =p.

that \P\ > | G |

1 / 2

T h e n O ( G ) > 1 , a n d thus
p

40

1. S y l o w

Theory

P r o o f . W e have \ G : P \ C o r o l l a r y 1.39.

|G| /|P|

< \ G \ , and thus O ( G ) > 1 by

p

W e m e n t i o n t h a t the conclusion of B r o d k e y ' s theorem definitely can fail if the Sylow p-subgroups are nonabelian. A counterexample is a certain group G of order 144 = 2 - 3 for which 0 ( G ) = 1. T a k i n g p = 2, we see that the conclusion of C o r o l l a r y 1.40 fails for this group, and thus the conclusion of B r o d k e y ' s theorem must also fail. O f course, the Sylow 2subgroups of G are necessarily nonabelian i n this case.
4 2 2

T h i s counterexample of order 144 can be constructed as follows. L e t E be elementary abelian of order 9. (In other words, E is the direct product of two cyclic groups of order 3.) T h e n E can be viewed as a 2-dimensional vector space over a field of order 3, and thus the full a u t o m o r p h i s m group of E is isomorphic to G L ( 2 , 3 ) , w h i c h has order 48. A Sylow 2-subgroup T of this a u t o m o r p h i s m group, therefore, has order 16, and of course, since T C A u t ( 7 ) , we see t h a t T acts on E and t h a t this action is faithful. U s i n g the "semidirect product" construction, w h i c h we discuss i n C h a p ter 3, we can b u i l d a group G containing (isomorphic copies of) E and T , where E < G = E T , a n d the conjugation action of T o n E w i t h i n G is identical to the original faithful action of T on E . N o w 0 ( G ) < G and E < G , and since 0 ( G ) is a 2-group and E is a 3-group, we see t h a t 0 ( G ) f l = l . It follows t h a t 0 ( G ) centralizes E , and thus 0 ( G ) is contained i n the kernel of the conjugation action of T on E . T h i s action is faithful, however, and we deduce t h a t 0 ( G ) = 1, and indeed we have a counterexample.
2 2 2 2 2 2

Problems

IF

1 F . 1 . I n the last paragraph of this section, we used the fact that i f M < G and N < G and M D N = 1, then M and N centralize each other. Prove this. H i n t . Consider elements of the form m - ^ m n , Note. Recall commutator element. N o t e i n more detail where m e M and n N .
l l

t h a t if x,y e G, then the element x ~ y - x y G is the of x and y . It is customary to write [ x , y ] to denote this t h a t [ x , y] = 1 i f and o n l y if xy = yx. W e s t u d y commutators i n C h a p t e r 4.
p

1F.2. L e t G be a group, and fix a prime p . Show t h a t if O ( G ) = 1, then there exist S , T e S y l ( G ) such t h a t Z ( 5 ) D Z ( T ) = 1.
p

1 F . 3 . L e t G = N P , where N < G , P < E S y l ( G ) and N n P = 1, a n d assume t h a t the conjugation action of P o n N is faithful. Show t h a t P acts faithfully on at least one orbit of this action.
p

1G

41

H i n t . I n fact, if x e N is chosen so t h a t | P n P \ is as s m a l l as possible, t h e n P acts faithfully on the P - o r b i t containing x .

x

1G Suppose t h a t A C G is an abelian subgroup, and t h a t the index n is relatively small. C a n we conclude t h a t G has a n o r m a l abelian TV such t h a t the index \ G : N \ is under control? Yes, certainly; t h a t \ G : c o r e ( , 4 ) | < n!, and of course, c o v e ( A ) is n o r m a l , and subgroup is contained i n A , it is abelian.
G G

= \ G: A \ subgroup we know since this

A b o u n d m u c h better t h a n \ G : N \ < n \ is available i f the abelian subgroup A happens to be a Sylow subgroup of G. A s s u m i n g that A is a Sylow p-subgroup, we can take N = O ( G ) , and then C o r o l l a r y 1.39 tells us t h a t \ G : N \ < n . Surprisingly, it is not necessary to assume t h a t A is a Sylow subgroup; there a l w a y s exists a n abelian n o r m a l subgroup N w i t h \ G : N \ < n . A l t h o u g h this remarkable theorem of A . C h e r m a k and A . Delgado does not depend on Sylow theory, it seems appropriate to include it here since it can be viewed as a generalization of C o r o l l a r y 1.39.
p 2 2

1.41. T h e o r e m ( C h e r m a k - D e l g a d o ) . L e t G be a finite g r o u p . T h e n G has a c h a r a c t e r i s t i c a b e l i a n s u b g r o u p N such t h a t \ G : N \ < \ G : A \ f o r every a b e l i a n s u b g r o u p A C G.

2

As we shall see, the key idea i n the proof of the C h e r m a k - D e l g a d o theorem is very elementary. B u t it is powerful, and it yields even more t h a n we stated i n T h e o r e m 1.41. T h e t r i c k is to define the integer m ( H ) = \ H \ \ C ( H ) \ for a r b i t r a r y subgroups H of a finite group G, a n d to consider subgroups J i C G where this C h e r m a k - D e l g a d o measure is m a x i m i z e d .
G G

We

begin w i t h a t r i v i a l observation.

1.42. L e m m a . L e t H C G, w h e r e G i s a finite g r o u p , a n d w r i t e C = C ( H ) . T h e n m ( H ) < m ( C ) a n d if e q u a l i t y o c c u r s , t h e n H = C ( C ) .

G G G G

P r o o f . Since H C C ( C ) , we see t h a t
G

m { C ) = |C||C (C)| > \ C \ \ H \ = m ( H ) ,

G G G

a n d the result follows.

N e x t , we consider two subgroups simultaneously. 1.43. L e m m a . L e t H a n d K be s u b g r o u p s D = H D K and J = ( H , K ) . Then m (H)m (K)

G G

of a

finite

g r o u p G, a n d

write

<

m (D)m (J)
G G G

a n d if e q u a l i t y h o l d s , t h e n J = H K a n d C ( D ) =

C (H)C {K).
G G

42

1. S y l o w

Theory

P r o o f . W r i t e C , C , C and C j to denote the centralizers i n G of H , K , D and J , respectively, and note that C j = C n C and that C , C C G . Then
H K D H K H K D

|J| > | M | =

and

Iwl and it follows t h a t (n\

m (D)
G

= \D\\CD\ > m

inli^

i^^MliM ^l r=rn
|J| \Cj\
D

rn (H)m (K) > m { J )

G G G

w h i c h establishes the inequality. If equality occurs, then \ J\ = \ H K \ and \C \ = \C CKI

D H

and thus J = H K and G

= C

as required.

W e m e n t i o n t h a t if H , K C G are subgroups, then the subgroup J = (7i, K ) is often called the j o i n of H and K . A l s o , a collection of subgroups of a group G is said to be a lattice of subgroups i f it is closed under intersections and joins. 1.44. T h e o r e m . G i v e n a finite g r o u p G, l e t C = C ( G ) be t h e c o l l e c t i o n of s u b g r o u p s of G f o r w h i c h t h e C h e r m a k - D e l g a d o m e a s u r e i s as l a r g e as possible. T h e n (a) C i s a l a t t i c e of s u b g r o u p s o f G . (b) I f H , K e (c) I f H e C , C, t h e n ( H , K ) = H K . then C ( H ) e C a n dC ( C ( H ) ) = H .
G G G

P r o o f . L e t m be the m a x i m u m of the C h e r m a k - D e l g a d o measures of the subgroups of G , and let H , K e C. T h e n m ( H ) = m = m ( K ) , and so b y L e m m a 1.43, we have m = m ( H ) m ( K ) < m ( D ) m ( J ) , where D = H n K and J = ( H , K ) . B u t m ( D ) < m and m ( J ) < m by the m a x i m a l i t y of m, and thus m ( D ) = m = m ( J ) , and hence D and J lie in C, as required. A l s o , since equality holds i n L e m m a 1.43, we k n o w t h a t H K = J , and this proves (b). Finally, statement (c) follows because the m a x i m a l i t y of m { H ) forces equality i n L e m m a 1.42.
G G 2 G G G G G G G G G

1.45. C o r o l l a r y . E v e r y finite g r o u p G c o n t a i n s a u n i q u e s u b g r o u p M , m i n i m a l w i t h t h e p r o p e r t y t h a t m ( M ) i s t h e m a x i m u m of t h e C h e r m a k - D e l g a d o measures of t h e s u b g r o u p s o f G . A l s o M i s a b e l i a n a n d M D Z ( G ) .

G

P r o o f . Since C = C { G ) is a lattice, the intersection of a l l of its members lies i n C, and this is the desired subgroup M . A l s o , since M e C, we know t h a t C ( M ) e C, and thus M C C ( M ) , and it follows t h a t M is abelian. F i n a l l y , M = C ( C ( M ) ) D Z ( G ) , and the proof is complete.
G G G G

Problems I G

43

We shall refer to the subgroup M of C o r o l l a r y 1.39 as the D e l g a d o subgroup of G. It is, of course, characteristic i n G. P r o o f of T h e o r e m 1.41. Since the acteristic and abelian, it suffices to abelian subgroups A C G. B y the m ( A ) = \ A \ \ C ( A ) \ > \ A \ , where abelian. T h e n
2 G G

Chermak-

C h e r m a k - D e l g a d o subgroup M is charshow t h a t \ G : M \ < \ G : A \ for a l l definition of M , we have m ( M ) > the last inequality holds because A is
2 G

\G A \ as required.

>

>

= \G M l

A n o t h e r a p p l i c a t i o n of the C h e r m a k - D e l g a d o subgroup is the following. 1.46. C o r o l l a r y . L e t H be a s u b g r o u p of a finite g r o u p G, a n d assume \ H \ \ C ( H ) \ > \G\. Then G is not a nonabelian simple group.
G

that

P r o o f . L e t M be the C h e r m a k - D e l g a d o subgroup of G and observe t h a t m ( M ) > m ( H ) > \ G \ . B u t the C h e r m a k - D e l g a d o measure of the identity subgroup is equal to | G | , and it follows t h a t M > 1. Since M is abelian and n o r m a l i n G, the result follows.
G G

Problems 1 G 1 G . 1 . L e t A Q G , where A is abelian, and assume t h a t there does not exist a characteristic abelian subgroup N of G such that \ G : N \ < \ G : A \ . Show t h a t A = C ( A ) is a member of the maximum-measure lattice C ( G ) and that \ G : Z { G ) \ = \ G : A \ .
2 G 2

1 G . 2 . L e t C ( G ) be the maximum-measure lattice as i n T h e o r e m 1.44, and suppose t h a t H e C { G ) and that H < G. Show t h a t there exists a n o r m a l subgroup M of G such t h a t H C M < G. H i n t . Observe t h a t a l l conjugates of H i n G lie i n C ( G ) and thus the product of any two of t h e m is a subgroup. If H is not contained i n a proper n o r m a l subgroup of G, show t h a t it is possible to write G = H K , where K < G and K contains a conjugate of H . Deduce a c o n t r a d i c t i o n from this. 1 G . 3 . L e t G be simple and suppose t h a t H C G and t h a t \ H \ \ C ( H ) \
G

\ G \ . Show t h a t H = 1 or H = G. H i n t . If G is nonabelian, show t h a t H C ( G ) . 1 G . 4 . L e t A C G , where A is abelian and G is nonabelian. Show t h a t there exists a n o r m a l abelian subgroup N of G such t h a t | G : / V | < | G : , 4 | .
2

Chapter 2

Subnormality

2A W e know t h a t i n general, n o r m a l i t y is not a transitive relation on the set of subgroups of a group. In other words, i f K < H and H < G, we cannot usually conclude that K < G. F o l l o w i n g W i e l a n d t , who developed m u c h of the theory t h a t we discuss i n this chapter, we "repair" this lack of t r a n s i t i v i t y b y defining a new relation. A subgroup S C G is said to be s u b n o r m a l i n G if there exist subgroups H i of G such t h a t S = H <Hi < <H = G,

and i n this s i t u a t i o n , we write S G. S u b n o r m a l i t y , therefore, is a transitive relation t h a t extends the n o t i o n of normality. W e stress that the chain of subgroups H i extending from S up to G is required to be finite, even if G is an infinite group. N o t e that a l t h o u g h we have not required t h a t the subgroups H i are a l l distinct, it is always possible to delete repeated groups and to assume that H < H < < H .
0 x r

O f course, if S G, there m a y be a number of different chains of subgroups { H i } t h a t realize the subnormality. T h e length of the shortest possible chain, or e q u i v a l e n t ^ the smallest possible integer r , is called the s u b n o r m a l d e p t h of S. T h u s the whole group G has d e p t h 0 i n G ; proper n o r m a l subgroups have depth 1, and s u b n o r m a l subgroups t h a t are not a c t u a l l y n o r m a l have depth exceeding 1. A l t h o u g h we are concerned almost exclusively w i t h finite groups i n this book, it should be noted t h a t part of the s n b n o r m a l i t y theory t h a t we present also works for infinite groups. Some of the theorems t h a t we prove by induct i o n o n the group order can also be proved by i n d u c t i o n on the s u b n o r m a l 45

46

2.

Subnormality

depth, and most of those results hold for infinite groups too. B u t the reader should be cautioned t h a t m a n y of the more interesting theorems about subn o r m a l i t y are s i m p l y false for general infinite groups. We 2.1. begin w i t h a s i t u a t i o n where subnormality arises naturally. L e t G be finite. T h e n G i s n i l p o t e n t if a n d o n l y if every

Lemma.

s u b g r o u p of G i s s u b n o r m a l . P r o o f . B y T h e o r e m 1.26, a finite group is nilpotent if and only if norm a l i z e s grow. I n other words, G is nilpotent if and only if H < N { H ) whenever H < G. N o w if H < G and H is subnormal, we can write H = H < < H = G, where r > 0, and we can assume t h a t H i > H . Since H = H < H i , w e have H < H i C N ( H ) , and this shows that if every subgroup of G is subnormal, then normalizers grow, and hence G is nilpotent.
G 0 r Q 0 G

Conversely, suppose t h a t G is nilpotent, so t h a t normalizers grow i n G. G i v e n H C G, we prove t h a t H G by i n d u c t i o n on the index \ G : H \ . If this index is 1, then H = G, and H is certainly s u b n o r m a l . Otherwise, H < G, and we have H < N ( H ) . T h e n \ G : N ( H ) \ < \ G : H \ , and so by the i n d u c t i v e hypothesis, N ( H ) G. T h e result now follows since H < N (H). U
G G G G

If we wanted to prove t h a t a subgroup S of G is s u b n o r m a l by using the definition, we w o u l d need to construct a chain of subgroups r u n n i n g from S up to G, where each of these subgroups is n o r m a l i n the next. I n the second half of the proof of L e m m a 2.1, we found such a chain b y being greedy. B y this, we mean t h a t i n going from H i to H , we went as far as possible: we took H i to be the full normalizer of H i i n G. B u t greediness does not always y i e l d success. In the s y m m e t r i c group S , for example, consider any subgroup S of order 2 i n the n o r m a l K l e i n subgroup K of order 4. (Recall t h a t the nonidentity elements of K are exactly the permutations i n S that are products of two disjoint 2-cycles.) Since K is abelian, we have S < K , and since K < S , we see t h a t S S . It is not h a r d to see, however, t h a t N ( 5 ) has order 8; it is a Sylow 2-subgroup of 5 . A l s o , since this S y l o w subgroup has index 3 and is not n o r m a l , it follows that it is its o w n normalizer. I n other words, we are stuck. W e can not continue to construct a s u b n o r m a l chain from N ( 5 ) to G, and so greediness does not work i n this case.
i + i l + 4 A 4 4 G 4 G

We pursue the connection between nilpotence and s u b n o r m a l i t y a bit further. R e c a l l t h a t the F i t t i n g subgroup F ( G ) of a finite group G is the unique largest n o r m a l nilpotent subgroup of G. T h e following result shows that not only does F ( G ) contain a l l n o r m a l nilpotent subgroups of G; it also contains a l l s u b n o r m a l nilpotent subgroups.

2A

47

2.2. T h e o r e m . L e t H C G, w h e r e G i s i f H i s n i l p o t e n t a n d s u b n o r m a l i n G.

finite.

T h e n H C F ( G ) if a n d

only

P r o o f . If H C F ( G ) , then of course is nilpotent since F ( G ) is nilpotent. A l s o , we see b y L e m m a 2.1 that iZ" F ( G ) < G , and thus H G , as wanted. Conversely, assume that H G and t h a t H is nilpotent. W e proceed by i n d u c t i o n o n |G| to show that H C F ( G ) . If H = G, t h e n G is nilpotent, and so H = G = F ( G ) , and we are done i n this case. W e can assume, therefore, t h a t H < G , and we let M be the penultimate t e r m i n a s u b n o r m a l chain of distinct subgroups c l i m b i n g from H to G . T h e n H M < G and M < G . B y the i n d u c t i v e hypothesis applied i n M , therefore, we have H C F ( M ) . B u t F ( M ) is characteristic i n M , w h i c h is n o r m a l i n G , and hence F ( M ) is a n o r m a l nilpotent subgroup of G . T h i s yields H C F ( M ) C F ( G ) , as required. W e leave nilpotent subgroups for a w h i l e to s t u d y s u b n o r m a l i t y more generally. W e begin w i t h a basic (and t r i v i a l ) fact. 2.3. L e m m a . L e t S G, w h e r e G i s a g r o u p , a n d suppose an arbitrary subgroup. Then S n K K . that K C G is

P r o o f . Since S is s u b n o r m a l i n G , there exist subgroups H i C G for 0 < i < r , where < H i for 0 < i < r , and where = 5 and H = G. Since < f i i , it follows that P i _ i n AT < H i n A", and so we have
0 r

5ni^

= p

n K < P i n A : ^ - - - ^ i i n A : = G n A : = A:,
r

and thus S (1 K

K , as required.

of a g r o u p G. Then

2.4. C o r o l l a r y . L e i 5 and T be s u b n o r m a l s u b g r o u p s SHTG.

P r o o f . W e have S n T T by L e m m a 2.3, and since T <M G , it follows by the t r a n s i t i v i t y of s u b n o r m a l i t y t h a t S n T G , as asserted. If 5 and T are n o r m a l subgroups of a group G , then, of course, their intersection S n T is also n o r m a l , and so too is their j o i n , ( 5 , T ) = ST. I n other words, the collection of n o r m a l subgroups of a n a r b i t r a r y group forms a sublattice of the full subgroup lattice. It is perhaps surprising, and it is certainly less t r i v i a l , t h a t the same conclusion holds for s u b n o r m a l subgroups of finite groups. W e have just seen t h a t the collection of s u b n o r m a l subgroups of G is closed under intersections, and t h a t was easy. T h e deeper result is that this collection is also closed under joins if |G| is finite. W e should mention, however, that there are counterexamples t h a t show that

48

2.

Subnormality

joins of s u b n o r m a l subgroups can fail to be s u b n o r m a l i n the general infinite case. 2.5. T h e o r e m . (S, T ) G. L e t G be a finite g r o u p , a n d suppose t h a t S,T G. Then

T h e r e are several k n o w n proofs of this remarkable theorem of W i e l a n d t , w h i c h was first established i n 1939. W i e l a n d t ' s original argument proceeded via double i n d u c t i o n directly from the definition of subnormality, and it is also valid for some infinite groups: those w h i c h satisfy the ascending chain c o n d i t i o n on s u b n o r m a l subgroups. O u r proof, on the other hand, relies on a later result of W i e l a n d t , and it makes sense o n l y i n the finite case. B u t it is less technical t h a n W i e l a n d t ' s , and it seems somewhat more conceptual. Before we prove T h e o r e m 2.5, we need to develop some more theory, i n c l u d i n g the following pretty result (of W i e l a n d t , of course). 2.6. T h e o r e m . L e t S G, w h e r e G i s a finite g r o u p , a n d l e t M m i n i m a l n o r m a l s u b g r o u p o f G . Then M C N (S).
G

be a

R e c a l l t h a t a m i n i m a l n o r m a l subgroup of G is a nonidentity n o r m a l subgroup M of G such t h a t the o n l y n o r m a l subgroup of G t h a t is properly contained i n M is the identity. F o r example, if G is simple, then G is a m i n i m a l n o r m a l subgroup of itself. To prove T h e o r e m 2.6, we must digress briefly to discuss the socle of a finite group G, w h i c h is the subgroup generated by a l l m i n i m a l n o r m a l subgroups of G, a n d w h i c h is denoted S o c ( G ) . ( O f course, the subgroup generated by a collection of n o r m a l subgroups is just the p r o d u c t of those subgroups.) T h e key observation here is t h a t since G is finite, every nonidentity n o r m a l subgroup N of G contains a m i n i m a l n o r m a l subgroup of G, and thus N n S o c ( G ) > 1. I n particular, if G > 1, then S o c ( G ) > 1. We r e m i n d the reader of a useful general fact, w h i c h is needed i n the proof of T h e o r e m 2.6. 2.7. L e m m a . L e t M a n d N be n o r m a l s u b g r o u p s of a g r o u p G, a n d M n N = 1. Then every element of M commutes w i t h every element assume of N .

P r o o f . L e t m M and n E N and consider the c o m m u t a t o r c = [ m , n ] , w h i c h , by definition, is the element m ^ n ^ m n . Since c = m m and M is n o r m a l , we see t h a t c G M . B u t also, c = ( n ) n , and thus c G N because TV is n o r m a l . T h e n c G M n N = 1, and so 1 = c = m m , a n d we deduce t h a t m = m. It follows that m and n commute, as asserted.
_ 1 n _ 1 m _ 1 n n

P r o o f of T h e o r e m 2.6. W e proceed by i n d u c t i o n on | G | . T h e r e is n o t h i n g to prove i f S = G, and so we can assume t h a t S < G. T h e n since S G, we

2A

49

can reason as we d i d previously to find a subgroup N < G w i t h S << N < G. ( S i m p l y take TV to be the penultimate t e r m of a s u b n o r m a l series w i t h distinct terms c l i m b i n g from S to G.) There are now two cases. If M n N = 1, t h e n by L e m m a 2.7, we have M C C ( N ) C C (S) C N ( S ) , as wanted. W e can suppose, therefore, t h a t 1 < M n N . T h e n since M n N C M , it follows from the fact t h a t M is m i n i m a l n o r m a l i n G that M C\ N = M , and we have M C N .
G G G

Since S N < G, the inductive hypothesis guarantees that every m i n i m a l n o r m a l subgroup of N normalizes S. N o w M < N , but we do not know t h a t M is m i n i m a l n o r m a l i n N , a n d so we cannot conclude directly t h a t M normalizes S. B u t we do k n o w t h a t Soc (TV) normalizes S, and so it w i l l suffice to show t h a t M C Soc (iV). F i r s t , observe t h a t M n Soc (iV) > 1 since 1 < M < N . A l s o , Soc (TV) is characteristic i n N , and hence it is n o r m a l i n G, a n d thus M n Soc (TV) < G . But 1 < M n Soc (iV) C M , and it follows from the fact t h a t M is m i n i m a l n o r m a l i n G t h a t M n Soc (TV) = M . T h e n M C Soc (TV) C N ( 5 ) , and the proof is complete.
G

F i n a l l y , we are ready to prove t h a t i n finite groups, joins of s u b n o r m a l subgroups are s u b n o r m a l . P r o o f of T h e o r e m 2.5. W e prove b y i n d u c t i o n on \ G \ t h a t if S,T i G, then (S, T ) G. Since we can certainly assume t h a t G > 1, we can choose a m i n i m a l n o r m a l subgroup M of G. L e t G = G / M , a n d observe t h a t the images S and T of S and T are s u b n o r m a l i n G. B u t \ G \ < \ G \ , and so the i n d u c t i v e hypothesis applies, a n d we conclude t h a t ( S , T ) G. Since "overbar" is a h o m o m o r p h i s m w i t h kernel M , it follows t h a t (S,T) and thus { S , T ) M G. = (S,T) = (S,T)M,

The correspondence theorem yields a bijection between the set of a l l subgroups of G a n d the set of a l l of those subgroups of G that contain M . F u r t h e r m o r e , this bijection preserves normality, a n d hence it also preserves subnormality, and we conclude from this t h a t ( S , T ) M G. B u t M is m i n i m a l n o r m a l i n G, and thus b y T h e o r e m 2.6, each of the subgroups S and T is n o r m a l i z e d by M . It follows t h a t M normalizes ( S , T ) , and we see t h a t ( S , T ) < ( S , T ) M G, and thus ( S , T ) is s u b n o r m a l i n G, as wanted. Two subgroups H and K o a group G are said to be p e r m u t a b l e if H K = K H . (Recall that this is equivalent to H K being a subgroup.) W e know t h a t each n o r m a l subgroup of G is permutable w i t h a l l subgroups,

50

2.

Subnormality

and it is of some interest to ask if the converse is true. Is a subgroup that is permutable w i t h a l l subgroups necessarily normal? T h e answer is "no", but it is true (in a finite group) that such a subgroup must be s u b n o r m a l . (A subgroup permutable w i t h a l l subgroups i n some group G is said to be q u a s i n o r m a l i n G.) A c t u a l l y , a result stronger t h a n "quasinormal implies s u b n o r m a l " is true. G i v e n S C G , it is enough to know that S is permutable w i t h a l l of its own conjugates to deduce t h a t S is subnormal; one need not assume t h a t S is permutable w i t h every subgroup. 2.8. SS
X

Theorem. = SS
X

L e t S C G, w h e r e G i s a G.

finite

g r o u p , a n d assume

that

for all x e G. Then S

Since it is certainly not obvious how to construct a s u b n o r m a l series from S to G i n the s i t u a t i o n of T h e o r e m 2.8, we must find some more subtle m e t h o d of proof. T r y i n g i n d u c t i o n on | G | , we see t h a t i f S C H < G, then the hypothesis on S is a u t o m a t i c a l l y satisfied i n H , and so S H by the i n d u c t i v e hypothesis. T o see what good t h a t might be, imagine what the s i t u a t i o n w o u l d be i f we were t r y i n g to prove t h a t S is n o r m a l i n G . We w o u l d then k n o w t h a t every proper subgroup of G t h a t contains S is contained i n N ( S ) , and so i f S is not actually n o r m a l i n G , then N ( S ) w o u l d be a m a x i m a l subgroup of G . I n fact, it w o u l d be the unique m a x i m a l subgroup containing S.
G G

B y the following "zipper l e m m a " of W i e l a n d t , a similar conclusion holds when we are t r y i n g to prove t h a t S is s u b n o r m a l , and this is the key to the proof of T h e o r e m 2.8. 2.9. T h e o r e m (Zipper L e m m a ) . Suppose that S C G, where G is a finite g r o u p , a n d assume t h a t S H f o r every p r o p e r s u b g r o u p H of G t h a t c o n t a i n s S. If S i s n o t s u b n o r m a l i n G, t h e n t h e r e i s a u n i q u e m a x i m a l s u b g r o u p of G t h a t c o n t a i n s S. A s we have seen, the analog of the zipper l e m m a is true (and t r i v i a l ) if "normal" replaces " s u b n o r m a l " , and i n that situation, the unique m a x i m a l subgroup of G that contains S is the normalizer of S. B u t i n general, there is no such t h i n g as a subnormalizer, and so we must work harder to prove T h e o r e m 2.9. W h a t we m e a n b y saying that subnormalizers do not exist is t h a t if S C G , there need not be a unique subgroup m a x i m a l w i t h the p r o p e r t y t h a t it contains S subnormally. F o r example, a Sylow 2-subgroup P of the simple group G = P S L ( 2 , 7) of order 168 is contained i n two m a x i m a l subgroups of G , each of order 24. It is not h a r d to check t h a t the subgroup Z ( P ) , w h i c h has order 2, is s u b n o r m a l i n each of these m a x i m a l subgroups, but of course, Z ( P ) cannot be s u b n o r m a l i n G because G is simple.

2A

51

Before we prove the zipper l e m m a , we demonstrate how it can be used by p r o v i n g T h e o r e m 2.8. W e need an easy p r e l i m i n a r y result. 2.10. L e m m a . L e t H C G, w h e r e G i s a n a r b i t r a r y g r o u p , a n d
X

suppose

that H H

= G. T h e n H = G.
x l

P r o o f . W r i t e x = uv, where u e H and v G H . T h e n xv~ have

= u a n d we

where the first equality holds because v 6 H u e H . Then G = H H

X

a n d the t h i r d holds because

= H H H a n d the proof is complete.

We also need a bit of notation. If H is an a r b i t r a r y subgroup of G, we write H = ( H \ x <E G ) . T h i s is the n o r m a l closure of H i n G, and we see that H < G because conjugation by elements of G s i m p l y permutes the generating subgroups of H . I n fact, H is the unique smallest n o r m a l subgroup of G containing H since i f I f C N < G, t h e n N contains a l l conjugates of H , and thus N contains the subgroup H generated by these conjugates.
G x G G G G

P r o o f of T h e o r e m 2.8. W o r k i n g by i n d u c t i o n on \ G \ , we can assume that S is s u b n o r m a l i n every subgroup H such t h a t S C H < G. A s s u m i n g t h a t S is not s u b n o r m a l i n G, we have S < G, a n d we derive a c o n t r a d i c t i o n . B y the zipper l e m m a , S is contained i n a unique m a x i m a l subgroup M of G. N o w consider a conjugate S of S. Since S < G, L e m m a 2.10 guarantees that SS < G, and since SS is a subgroup because SS = S S by hypothesis, we conclude that SS is contained i n some m a x i m a l subgroup of G. B u t the only m a x i m a l subgroup c o n t a i n i n g S is M , and so we must have SS C M . T h e n S C M for a l l elements x e G.
x X X X X X X x

Now consider the n o r m a l closure S = (S | x G ) . B y the result of the previous paragraph, we know that S C M , and thus S < G. It follows t h a t S S < G, and thus S G. T h i s is a c o n t r a d i c t i o n , and the proof is complete.
G G G

We

now prove the W i e l a n d t zipper l e m m a .

P r o o f of T h e o r e m 2.9. W o r k by i n d u c t i o n on \ G : 5 | . Since we are as. s u m i n g t h a t S is not s u b n o r m a l i n G, we certainly have S < G, and thus the case \ G : S\ = 1 of the theorem is vacuously satisfied a n d our i n d u c t i o n starts. A s S is not n o r m a l , we have N ( S ) < G, and thus N { S ) C M for some m a x i m a l subgroup M of G. O f course S C M , and our task is to show
G G

52

2.

Subnormality

t h a t M is the o n l y m a x i m a l subgroup of G that contains S. Suppose then, t h a t also S C i f , where i f is m a x i m a l i n G. W e work to show that K M . Now S Q K < G, and so by hypothesis, 5 i f . Suppose that i n fact S < K . T h e n i f C N ( S ) C M , and since i f is m a x i m a l i n G, it follows t h a t i f = M , as wanted. W e can suppose, therefore, t h a t S is not n o r m a l in i f .
G

Sinee S

i f , there is some shortest s u b n o r m a l series S = Ho < H i o o H = if,

and we see t h a t r > 2. A l s o , S is not n o r m a l i n H since otherwise we could delete H i to o b t a i n a shorter s u b n o r m a l series for S i n i f . L e t x G I f w i t h S + S a n d w r i t e T = (5, S*). N o t e that T C i f since 5 C i f and x e i f . A l s o , S C ( H i ) = H i C N G ( 5 ) , and thus T C N G ( 5 ) C M . Furthermore, we see t h a t S < T < G.
2 2 x x x

Observe that S satisfies the hypotheses imposed o n S of the theorem. ( T h i s , of course, is because conjugation a u t o m o r p h i s m of G.) W e c l a i m t h a t the subgroup T = (S, these hypotheses. Specifically, we need to show that if T T oo H and that T is not s u b n o r m a l i n G.

i n the statement by x defines a n S ) also satisfies C H < G, then

x

F i r s t , if T C H < G, then clearly S Q H , a n d thus S oo H , and similarly, oo H . T h e n T is the j o i n of two s u b n o r m a l subgroups of H , a n d we conclude by T h e o r e m 2.5 that r oo H as wanted. A l s o , S o T , a n d so if T oo G, it w o u l d follow that S oo G, w h i c h is not the case. T h u s T is not s u b n o r m a l , as claimed. Now T > S since S is contained i n T, but it is not contained i n S. T h e n | G : T\ < \ G : 5|, a n d so by the inductive hypothesis, T is contained i n a unique m a x i m a l subgroup of G. O n the other hand, we k n o w that T C M a n d r C i f , a n d thus M = i f . T h e proof is now complete.
x

Perhaps we should e x p l a i n the name "zipper l e m m a " . T h e key idea of the proof of T h e o r e m 2.9 is t h a t if S is contained i n two different m a x i m a l subgroups, then some larger subgroup T is also contained i n two different m a x i m a l subgroups. R e p e a t i n g this argument, we w o u l d get a still larger subgroup contained i n two different m a x i m a l subgroups, and so on. B u t this cannot go on indefinitely i n a finite group. A s we c l i m b higher, the two m a x i m a l subgroups are forced to be the same. T h i s is analogous to z i p p i n g up a n open zipper. A s we p u l l up on the zipper p u l l , the two top parts of the zipper are forced together. We conclude this section w i t h another a p p l i c a t i o n of the zipper l e m m a .

Problems 2 A

53

2.11. T h e o r e m . L e t A be a n a b e l i a n s u b g r o u p of a finite g r o u p G, a n d assume t h a t f o r every s u b g r o u p H w i t h A C H C G, we have \ H : A \ < \H:Z{H)\. T h e n A C F { G ) .

2

The c o n d i t i o n that \ H : A \ < \ H : Z ( H ) \ for a l l subgroups H such t h a t A C H C G m a y seem somewhat artificial, but i n fact, it arises n a t u r a l l y i n character theory. If A is abelian a n d some character of A induces irreducibly to G, t h e n this c o n d i t i o n is a u t o m a t i c a l l y satisfied. A l s o , we observe t h a t if A is abelian a n d satisfies this c o n d i t i o n , t h e n A = C ( A ) . T o see this, take H = C ( A ) a n d note that \ H : Z { H ) \ < \ H : A \ , a n d so the assumed inequality forces \ H : A \ = 1.
G G

P r o o f of T h e o r e m 2.11. W o r k i n g by i n d u c t i o n o n | G | , we can assume t h a t A C F { H ) whenever A C H < G, a n d thus A P . A l s o , by T h e o r e m 2.2, we are done if A G , a n d so we can assume that A is not s u b n o r m a l i n G, a n d the zipper l e m m a applies. W e conclude, therefore, t h a t there is a unique m a x i m a l subgroup M c o n t a i n i n g A . Now let g <E G. If ( A , A ) < G, t h e n m a x i m a l subgroup of G, a n d hence { A , A ) C If this happens for a l l g G, t h e n A C M , A < M A < G, w h i c h is a c o n t r a d i c t i o n . W e some element g G.
9 G g 9

{ A , A ) is contained i n some M , a n d i n p a r t i c u l a r , ,4 C M . a n d thus ^4 < G a n d we have conclude that G = ( 4 , 4 ) for

G 9

Since A a n d

are abelian, we have A r \ A |G| >

C Z ( G ) , a n d it follows t h a t

= JJ > ' '

\AC\A9\ ~
2 2

\Z{G)\'

where the strict inequality follows by L e m m a 2.10 since A < G. M u l t i p l y i n g b o t h sides by | G | / | A |, , , we conclude t h a t \ G : A \ > \ |G G : : Z ( G ) | , contrary to the hypothesis. '
2

Problems 2A 2 A . 1 . L e t 7t be a set of prime numbers, a n d recall t h a t O (G) is the unique largest n o r m a l rr -subgroup of G . Show t h a t O ( G ) contains every s u b n o r m a l rr -subgroup of G . C o n c l u d e that the subgroup generated b y two s u b n o r m a l 7r-subgroups of G is itself a vr-subgroup.
f f w

2 A . 2 . A g a i n let vr be a set of prime numbers, a n d recall t h a t 0 * ( G ) is the unique smallest n o r m a l subgroup of G whose factor group is a rr -group. If K G a n d | G : K \ is a rr -number, show t h a t i f D 0 * ( G ) .

54

2.

Subnormality

2 A . 3 . L e t H , i f C G a n d suppose that \ G : H \ and \K\ are relatively prime. (a) If H (b) If i f G, show K Q H .

G, show i f C H .

2 A . 4 . L e t S C. G, where G is a finite group and S is simple, and suppose t h a t S H = H S for a l l s u b n o r m a l subgroups H of G. Show that 5 C N ( i / ) for a l l I f G .
G

N o t e . T h e intersection of the normalizers of a l l s u b n o r m a l subgroups of a group G is the W i e l a n d t s u b g r o u p of G . 2 A . 5 . L e t X be any collection of m i n i m a l n o r m a l subgroups of G , and let N = Yi x . (a) Show t h a t N is the direct product of some of the members of X . (b) Show that every m i n i m a l n o r m a l subgroup of N is simple. (c) Show that TV is a direct product of simple groups. H i n t . F o r (b), show t h a t Soc (TV) = N . N o t e . T h i s p r o b l e m shows that m i n i m a l n o r m a l subgroups and socles of finite groups are direct products of simple groups. 2 A . 6 . I n this s i t u a t i o n of the previous problem, show t h a t every nonabelian n o r m a l subgroup of G contained i n N contains a member of X . 2A.7. Let S G , where S is nonabelian and simple. Show that S
G

is a

m i n i m a l n o r m a l subgroup of G . H i n t . W o r k by i n d u c t i o n on \ G \ to conclude that S C Soc(Jf) whenever S C H . Deduce that each conjugate of S i n G is a m i n i m a l n o r m a l subgroup of S . T h e n a p p l y the previous p r o b l e m to the group S , where X is the set of a l l G-conjugates of S.
G G

2 A . 8 . L e t S a n d T be different nonabelian s u b n o r m a l simple subgroups of G . Show that S and T commute elementwise. 2 A . 9 . L e t HG and assume that H = 0 * { H ) , Show that C v ( G ) normalizes H .
n

where vr is a set of primes.

H i n t . It is no loss to assume that G = H O ( G ) . H = 0*(H).

Show i n this case t h a t

2 A . 1 0 . W e say that subgroups H , i f C G are strongly conjugate if they are conjugate i n the group ( H , i f ) . Show that H G if and o n l y if the o n l y subgroup of G that is strongly conjugate to H is H itself.

2B

55

2B In this section, we present a useful result t h a t is essentially due to R . B a e r , and w h i c h we derive from the W i e l a n d t zipper l e m m a . Versions of Baer's theorem have also been proved b y M . S u z u k i a n d by J . A l p e r i n a n d R . Lyons. 2.12. T h e o r e m ( B a e r ) . L e t H be a s u b g r o u p
x

of a

finite

g r o u p G.

Then

f f C F ( G ) if a n d o n l y if (ff, H ) i s n i l p o t e n t f o r a l l x G. P r o o f . O n e direction is t r i v i a l . If f f C F ( G ) , t h e n also H C F ( G ) since F ( G ) < G. T h e n ( H , H ) is a subgroup of the nilpotent group F ( G ) , a n d hence it is nilpotent.
X x

Conversely, assume that (ff, H ) is nilpotent for a l l x G G . I n particular, f f is nilpotent, a n d so t o prove that f f C F ( G ) , it suffices b y T h e o r e m 2.2 to prove t h a t f f G . W e establish this b y i n d u c t i o n o n | G | . A s s u m e that f f is not s u b n o r m a l i n G , a n d observe t h a t i f f f is contained i n a proper subgroup f f , then ( i f , H ) is nilpotent for a l l x K , a n d thus f f AT by the inductive hypothesis. T h e zipper l e m m a now applies, a n d we deduce that f f is contained i n a unique m a x i m a l subgroup M .
x

Now consider a conjugate H of f f . If ( i f , H ) = G , t h e n G is nilpotent, a n d hence f f G , w h i c h is not the case. T h e n ( i f , f f ) < G , a n d so this subgroup must be contained i n some m a x i m a l subgroup of G. B u t (ff, H ) contains f f , a n d the only m a x i m a l subgroup of G t h a t contains f f is M . It follows t h a t (ff, f f ) C M , a n d we conclude t h a t M contains a l l conjugates of f f i n G , a n d so H C M . T h e n o r m a l closure i f is thus proper i n G , a n d we have f f C H < G. T h e n , H H < G , a n d so i f G . T h i s c o n t r a d i c t i o n shows t h a t f f G , as required.
x x x G G G G

The following is a useful consequence of Baer's theorem. T h e o r e m 2.13 is the key t o M a t s u y a m a ' s extension to groups of even order of G o l d s c h m i d t ' s non-character proof of the B u r n s i d e pV-theorem. ( G o l d s c h m i d t required that b o t h primes were o d d , a n d M a t s u y a m a removed that restriction.) R e c a l l that a n involution i n a group G is any element of order 2. 2.13. T h e o r e m . L e t t be a n i n v o l u t i o n i n a t h a t t # 0 ( G ) . T h e n t h e r e exists a n element such t h a t x = x ~ .
2 l l 2

finite g r o u p G, a n d x G of o d d p r i m e

assume order

Since 0 ( G ) is the unique largest n o r m a l 2-subgroup of G , the c o n d i t i o n that t 0 0 ( G ) says t h a t t is not contained i n any n o r m a l 2-subgroup of G .
2

In order to prove T h e o r e m 2.13, we need t o digress briefly. A group D is d i h e d r a l i f it contains a n o n t r i v i a l cyclic subgroup C of index 2 such that every element of D - C is a n i n v o l u t i o n . O f course, \ D \ = 2 | G | , a n d so d i h e d r a l groups have even order at least 4, a n d it is not h a r d t o show

56

2.

Subnormality

t h a t a d i h e d r a l group is determined up to i s o m o r p h i s m b y its order. G r o u p theorists usually refer to the d i h e d r a l group of order 2 n as D , and that is the n o t a t i o n we use here. ( B u t this n o t a t i o n is not completely standard, and this group is often referred to as D . ) T h e dihedral group D is isomorphic to the K l e i n group C x C , a n d except for this degenerate case, d i h e d r a l groups are always nonabelian. It is easy to see that D exists for n > 2: it is the group of r o t a t i o n a l symmetries i n 3-space of a regular n-gon. (In C h a p t e r 3, we shall see how to construct dihedral groups as semidirect products, thereby establishing their existence w i t h o u t relying on geometric reasoning.)
2 n n 4 2 2 2 n

The

following presents some of the basic facts about d i h e d r a l groups.

2.14. L e m m a . L e t D be a g r o u p . (a) Suppose t h a t C = (c) i s a n o n t r i v i a l c y c l i c s u b g r o u p of D , and that t < E D - C i s a n i n v o l u t i o n . T h e n every D - C i s a n i n v o l u t i o n if a n d o n l y if c = c~ . I n t h i s d i h e d r a l , a n d x = x ' f o r a l l x C a n d y e D - C. g e n e r a t e d by t h e d i s t i n c t i n v o l u t i o n s c t a n d t.

l l y 1

index 2 i n element of case, D i s Also, D is

(b) Suppose t h a t D i s g e n e r a t e d by d i s t i n c t i n v o l u t i o n s s a n d t. T h e n the c y c l i c s u b g r o u p C = (st) i s n o n t r i v i a l , a n d i t fails t o c o n t a i n s a n d t. A l s o \ D : C\ = 2, a n d t i n v e r t s t h e g e n e r a t o r st of C. I n particular, w e are i nthe above situation, a n d D i s dihedral. P r o o f . F o r (a), observe t h a t D - C is exactly the coset C t . If a e C then ( a t ) = a t a t = a a \ a n d thus a} = a ' i f and only if ( a t ) = 1. In particular, if every element of D - C is an involution, then ( c t ) = 1, and so c = c " . Conversely, suppose that c* = c " . Since every element of C is a power of c, we have a* = a " for a l l a C, a n d hence ( a t ) = 1, and every element of D - C = C t is an involution. I n other words, D is dihedral. Let x E C a n d y D - C. T h e n y = a t for some element a G C, a n d we have x = x = x * = x ~ . A l s o , c t is a n involution, a n d the group ( c t , t ) properly contains C = (c), a n d thus it is the whole group D . F i n a l l y , c t a n d t are distinct because c ^ 1.
2 1 2 2 1 1 1 1 2 y a t l

Now assume the s i t u a t i o n of (b) a n d observe that C is n o n t r i v i a l since s + t. W e have (si)* = t ( s t ) t = ts = ( s t ) - \ and thus t inverts the generator st of C, as wanted. T h e n C * = ( s t ) = ((si)') = ( ( s i ) " ) = C, and so t e N ( C ) . It follows t h a t C ( t ) is a subgroup containing b o t h i and ( s i ) i = s, and thus C ( i ) is a l l of D . Because st < E C, we see t h a t i f either s or i lies in C, then b o t h s and t lie i n C, and this is impossible since a cyclic group can c o n t a i n at most one involution. T h u s C contains neither s or t, and i n particular, C n (i) = 1. Since D = C ( t ) , we have \ D : C\ = | ( i ) | = 2, and this completes the proof of (b).
1 1 D

Problems

2B

57

The fact t h a t two elements of order 2 i n a n a r b i t r a r y group always generate a d i h e d r a l subgroup does not seem to generalize. T h e r e appears to be almost n o t h i n g that can be said about the structure of a subgroup generated by two elements of given orders m > 1 a n d n > 1 i n any case other t h a n m = 2 = n . P r o o f of T h e o r e m 2.13. L e t T = ( t ) . If T C F ( G ) , t h e n T is contained in the unique Sylow 2-subgroup of F ( G ) , w h i c h is 0 ( G ) , contrary to hypothesis. B y B a e r ' s theorem (2.12) therefore, there exists g G such t h a t ( T , T ) is not nilpotent. In particular, ( t , t ) is not a 2-group.
2 9 9

B y L e m m a 2.14, it follows that D = ( t , t ) is d i h e d r a l of order 2 | G | , where C is a cyclic subgroup whose elements are inverted by t. Since D is not a 2-group, neither is G , and thus there exists a n element x C of o d d prime order. Since x = a T , the proof is complete.
t 1

Problems 2B 2 B . 1 . L e t H C G and assume that for each element g e G , either ( H , is nilpotent or H H = H H . Show that H G .
9 9

H )

2 B . 2 . L e t L> be the dihedral group of order 2 n . (a) If n is o d d , show that D contains e x a c t l y n involutions, and t h a t they a l l lie i n a single conjugacy class. (b) If n is even, show that D contains e x a c t l y n + 1 involutions, a n d these lie i n exactly three conjugacy classes, w i t h sizes 1, n / 2 and n/2, respectively. 2 B . 3 . L e t s a n d t be involutions i n a group G . If s a n d t are not conjugate in G , show t h a t there exists an involution, z e G, different from s a n d t, a n d c o m m u t i n g w i t h b o t h of t h e m . 2 B . 4 . Suppose that G has more t h a n one Sylow 2-subgroup and that every two distinct Sylow 2-subgroups of G intersect t r i v i a l l y . Show t h a t G contains exactly one conjugacy class of involutions. 2B.5. Let B Q G , (1) G - B b e B . (2) G-B consists entirely of involutions.
l

w i t h | G : B \ = 2. Show t h a t the following are equivalent.

1

contains an involution t such t h a t 6* = 6 "

for a l l elements

(3) E v e r y element t of G - B is an i n v o l u t i o n such t h a t 6* = b~ for a l l elements b G B . Show also t h a t if these conditions hold, then B must be abelian.

58

2.

Subnormality

N o t e . In this situation, G is said to be generalized dihedral. 2 B . 6 . Show that G has a n o r m a l Sylow p-subgroup if and only i f every subgroup of the form ( x . y ) has a n o r m a l Sylow p-subgroup, where x and y are conjugate elements of G h a v i n g p-power order. 2C A l t h o u g h there really is no connection w i t h subnormality, we digress i n this section to discuss the terms "local" and "global" as they are used i n finite group theory. ( A n d then we w i l l present an a p p l i c a t i o n of T h e o r e m 2.13 i n this context.) B y definition, a subgroup i f of a group G is p -local, where p is prime, if i f is of the form i f = N ( P ) , where P is some nonidentity p-subgroup of G. M o r e generally, a subgroup is local if it is p-local for some prime p. T h e idea here is that i n some sense, the local subgroups are those t h a t are easiest to find: choose a Sylow subgroup, pick a n o n t r i v i a l subgroup of it, and then take the normalizer.
G

For example, suppose we want to prove that up to isomorphism, the only simple group of order 60 is the alternating group A . T o establish this, it suffices to show that an a r b i t r a r y simple group G of order 60 must have a subgroup H of index 5. T h e n the action of G on the set of right cosets of H yields an i s o m o r p h i s m from G into A , and because | G | = 60, this m a p must be surjective, and G = A $ , as required.
5 5

G i v e n a simple group G of order 60, we can find the required subgroup ff, as follows. T h e number n of Sylow 2-subgroups of G must be either 5 or 15. If n = 5, take H to be the normalizer of a Sylow 2-subgroup of G. If n = 15, then since 1 5 - 1 is not divisible by 4, it follows by T h e o r e m 1.16 that there exist distinct Sylow 2-subgroups S and T t h a t intersect nontrivially. Let D = S n T > 1 and H = N ( D ) . T h e n f f contains b o t h S and T, and thus | i f | is a proper m u l t i p l e of 4. Since G is simple, D is not n o r m a l , and so i f < G. A l s o , \ G : i f | + 3 by the n!-theorem, a n d the only r e m a i n i n g possibility is t h a t i f has index 5, as wanted.
2 2 2 G

In b o t h cases, where n = 5 and where n = 15, we found a 2-local subgroup of index 5, and as we explained, this shows that G A . (Actually, since n ( A ) = 5, we see that the case n = 15 does not occur, but nevertheless, it h a d to be considered.)
2 2 5 2 5 2

Now consider the analogous p r o b l e m for A . It is true t h a t every simple group of order 360 is isomorphic to A , but this is considerably more difficult to prove. It w o u l d suffice to find a subgroup of index 6 i n a n a r b i t r a r y simple group of order 360, but this is h a r d to do because no l o c a l subgroup of A has index 6, and so no argument such as we used for A can possibly work. In fact, a l l subgroups of index 6 i n A are isomorphic to A , and hence
6 6 6 5 6 5

2C

59

they are simple. N o subgroup of index 6, therefore, can be the normalizer of a n y t h i n g i n a simple group of order 360. A l t h o u g h it is possible, using character theory, to show that a simple group of order 360 really does have a subgroup of index 6, the usual proof that up to i s o m o r p h i s m , there is a unique simple group of order 360 proceeds b y showing t h a t every such group is isomorphic to P S L { 2 , 9). ( O f course, it follows that A a n d P S L { 2 , 9) are isomorphic.)
6

Properties of groups that do not d i r e c t l y refer to l o c a l subgroups are often referred to as global properties. W h a t is remarkable, is how often it happens that global i n f o r m a t i o n can be deduced from appropriate d a t a about local subgroups. F o r example, here is a "local-to-global" result t h a t is not t e r r i b l y difficult to prove. 2.15. T h e o r e m . Suppose t h a t f o r every of a finite g r o u p G has a n o r m a l Sylow Sylow 2-subgroup. o d d p r i m e p , every 2 - s u b g r o u p . Then p-local subgroup G has a n o r m a l

If a group G has a n o r m a l Sylow 2-subgroup, it is easy to see that every subgroup of G also has a n o r m a l Sylow 2-subgroup. T h e o r e m 2.15 can thus be viewed as a strong converse to this remark. Its proof relies on T h e o r e m 2.13 together w i t h the general observation t h a t local subgroups of h o m o m o r p h i c images are images of l o c a l subgroups. (We state this a b i t more precisely, as follows.) 2.16. L e m m a . L e t N < G, a n d w r i t e G = G/N, u s i n g t h e s t a n d a r d " b a r convention", where overbar is the canonical h o m o m o r p h i s m G ^ G . Then f o r a l l p r i m e s p , every p - l o c a l s u b g r o u p of G has t h e f o r m L , w h e r e L i s some p - l o c a l s u b g r o u p of G. P r o o f . B y the correspondence theorem, every subgroup of G has H , for some subgroup H of G w i t h H D N . N o w let N C M C M is p-local i n G. A l t h o u g h M m a y not be p - l o c a l i n G, we w i l l the proof by showing t h a t M = L for some p - l o c a l subgroup L of the form G, where complete G.

Since M is p-locat, it is the n o r m a l i z e r j n G of some n o n t r i v i a l psubgroup, w h i c h we can write m the form U, where N < U C G a n d \ U : A | is a power of p. T h e n M = N ^(C7), a n d since U 2 N , it follows that M = N Q ( U ) . Let P G Sylp(C/), a n d let L = N ( P ) . N o w U = N P because N a n d P have coprime indices i n U, a n d since N < U, it follows t h a t P > 1, a n d thus L is p-local i n G. Since /^normalizes b o t h N a n d P , we have L C N ( N P ) = N ( U ) = M , and thus I C M . T o o b t a i n the reverse containment, observe that since U < M and P G S y l ( [ / ) , the F r a t t i n i argument yields
G G G p

= U N

( P )

= N P N M ( P )

= N N

( P )

C N L .

60

2.

Subnormality

T h e n M C N L = L , where the equality holds because N is the kernel of the overbar h o m o m o r p h i s m . It follows t h a t M = L , as wanted. P r o o f of T h e o r e m 2.15. F i r s t , we suppose t h a t 0 ( G ) = 1, and we show in this case t h a t \ G \ is o d d . If G has even order, we can choose an i n v o l u t i o n t e G, a n d we observe t h a t t & 0 ( G ) . T h e n by T h e o r e m 2.13, there exists an element x of o d d p r i m e order p such that x = x ~ . L e t X = ( x ) , and note that t G N { X ) a n d t h a t N ( A ) is p-local. B y hypothesis, N ( A ) has a n o r m a l Sylow 2-subgroup S, w h i c h necessarily contains t. A s X and S are n o r m a l i n N ( A ) a n d XnS = 1, it follows by L e m m a 2.7 that X a n d 5 centralize each other, a n d i n particular, t centralizes x . T h u s x = x = x " , and this is a c o n t r a d i c t i o n since x has order p > 2. T h i s shows that \ G \ is odd, as wanted.
2 2 t l G G G G t 1

In the general case, let N = O ^ G ) , and observe that by the previous l e m m a , the p-local subgroups of G = G / N are h o m o m o r p h i c images of p-local subgroups of G for a l l o d d primes p . B y hypothesis, the p-local subgroups of G have n o r m a l Sylow 2-subgroups, and thus the same is true for their h o m o m o r p h i c images, a n d it follows t h a t G satisfies the hypotheses of the theorem. B u t 0 ( G ) is t r i v i a l , and by the first part of the proof it follows that that \ G \ is o d d . T h e n N = 0 ( G ) is a n o r m a l Sylow 2-subgroup of G , a n d the proof is complete.
2 2

We close this section w i t h a l e m m a that w i l l be needed later. A l t h o u g h this result has n o t h i n g to do w i t h subnormality, it is related to L e m m a 2.16, a n d so it seems appropriate to present it here. T h e assertion of L e m m a 2.16 is that every p-local subgroup of a h o m o m o r p h i c image of G is the image of a p-local subgroup of G. T h i s suggests the question of whether or not the image of a p - l o c a l subgroup must be p-local. T h e answer is "not always". F o r example, if O ( G ) is n o n t r i v i a l , then G is p-local i n itself, but this cannot be true for G / O ( G ) , w h i c h has no nonidentity n o r m a l p-subgroup. B u t p-local subgroups do m a p to p-local subgroups if the kernel of the h o m o m o r p h i s m is a p'-group.
p p

2.17.

L e m m a . L e t G be a

finite

group,

a n d l e t N < G.

Write G =

G/N, if L i s

a n d l e t p be a p n m e t h a t does of G, t h e n P i s n o n t r i v i a l ,

not divide \N\.

If P i s a n o n t r i v i a l
G

p-subgroup

and N^(P) = N ( P ) .

I nparticular,

p - l o c a l i n G, t h e n L i s p - l o c a l i n G. P r o o f . Since the p-group P is n o n t r i v i a l , and the order of N is not divisible by p, we h a v e _ P % N , and thus P is n o n t r i v i a l . W r i t e L = N ( P ) , a n d observe t h a t P < L because overbar is a h o m o m o r p h i s m . T h u s I C I % ( P ) a n d it suffices to prove the reverse containment.
G

2 D

61

B y the correspondence theorem, every subgroup of G has the form X for some (unique) jsubgroup X C G w i t h X D TV, a n d so i n p a r t i c u l a r , we can w r i t e I % ( P ) = M , where TV C M C G. N o w P ~ N = P < M , and thus P / V < M by the correspondence theorem. A l s o , P S y l ( P / V ) since p does not d i v i d e \ N \ , a n d thus by the F r a t t i n i argument, w e j i a v e M = N ( P ) ( P / V ) = N ( P ) N C N ( P ) / V = L / V . T h i s yields N ^ ( P ) = M C L , as wanted. T h e last statement should now be clear.
p M M G

Problems 2C 2C.1. A finite group G is said to be a n N - g r o u p i f every local subgroup is solvable. (a) Show that every proper h o m o m o r p h i c image of a n N - g r o u p is solvable. (b) L e t G be a nonsolvable N - g r o u p . Show t h a t G has a unique m i n i m a l n o r m a l subgroup S, a n d t h a t S is n o n a b e l i a n simple. H i n t . T h e F r a t t i n i argument is relevant. A l s o , recall t h a t subgroups and h o m o m o r p h i c images of solvable groups are solvable and that i f K < L w i t h K a n d L / K b o t h solvable, then L is solvable. N o t e . A major step leading toward the classification of finite simple groups was J . T h o m p s o n ' s classification of nonsolvable N - g r o u p s . A nonabelian finite simple group is m i n i m a l simple i f every proper subgroup is solvable, and since m i n i m a l simple groups are clearly N - g r o u p s , T h o m p s o n ' s work provided a classification of a l l m i n i m a l simple groups.

2D N e x t , we present a n amazing result of Zenkov. It is another a p p l i c a t i o n of Baer's theorem. 2.18. T h e o r e m (Zenkov). L e t A a n d B be a b e l i a n s u b g r o u p s of a f i n i t e g r o u p G, a n d l e t M be a m i n i m a l member of t h e set { A f l B \g e G } . Then M C F(G).
9

The a s s u m p t i o n i n T h e o r e m 2.18 t h a t M is a " m i n i m a l member" of some set of subgroups means, as usual, t h a t no member of the set is p r o p e r l y contained i n M . O f course, if M is chosen to have m i n i m a l order among members of the set, then M w i l l necessarily be m i n i m a l i n this sense, but not conversely since i n general, a m i n i m a l member of some collection of subgroups need not have m i n i m a l order.

62

2.

Subnormality

Suppose that G has an abelian Sylow p-subgroup P . If we apply Zenkov's theorem w i t h A = P = B , we deduce that P n F C F ( G ) for some element g E G, and thus P n F C O ( G ) , w h i c h is the unique Sylow p-subgroup of F ( G ) . W e know, however, that i n general, O ( G ) is contained i n every Sylow p-subgroup of G , and it follows i n this case that P n F = O ( G ) . In other words, if G has abelian Sylow p-subgroups, then O ( G ) is the intersection of two of t h e m . T h i s is exactly B r o d k e y ' s theorem, w h i c h appeared i n the previous chapter as T h e o r e m 1.37. W e can view Zenkov's theorem, therefore, as a far-reaching generalization of B r o d k e y ' s result.
p p p p

P r o o f of T h e o r e m 2.18. T h e set { A n B \ g e G } is unchanged i f we replace B by an a r b i t r a r y G-conjugate B , and thus it is no loss to assume that M = A n B . W e show by i n d u c t i o n on | G | that M C F ( G ) . F i r s t , suppose that G = ( A , B ) for some element g G . Since A and B are abelian, we have A n B C Z ( G ) , and thus A n B = ( A nB ) ' C P . It follows t h a t A n B C A n B = M , and by the m i n i m a l i t y of M we have M = A n B C Z ( G ) C F ( G ) , as required.
9 9 9 9 9 9 9 9 1 9 9

We can now assume t h a t ( A , B ) < G for a l l g G G . T o show that M C F ( G ) , it is enough to prove for all primes p t h a t a full Sylow psubgroup P of M is contained i n F ( G ) . (This suffices, of course, because M is generated by its Sylow subgroups.) B y Baer's theorem (2.12), therefore, it is enough to show that ( P , P ) is nilpotent for a l l elements g G .
9

F i x g e G and let H = ( A , B ) , so that H < G. W r i t e C = B n i f , and observe that i f /* e i f , then ^ n G = , 4 n ( P n P ) = A n P n P = A n P \ where the final equality holds because A C P . In particular, M = i n B = ^4 n G is m i n i m a l i n the set { A n C \he H } . B y the inductive hypothesis applied i n P , and w i t h C replacing B , we conclude t h a t P C M C F ( P ) , and thus P C O ( P ) since O ( P ) is the unique Sylow p-subgroup of F ( P ) . A l s o , P C B C H , and thus P normalizes O ( i f ) and O ( H ) P is a p-group containing ( P , P ) . I n particular, ( P , P ) is a p-group, and hence it is nilpotent, as required.
9 / l / l / l h p p 9 9 9 9 p p 9 9

2.19. C o r o l l a r y . L e t A C G, w h e r e A i s a b e h a n a n d G i s a n o n t r w i a l g r o u p , a n d assume that \A\ > \ G : A \ . T h e n A n F ( G ) > 1.

9 2

finite

P r o o f . If g G, we compute that \ A \ \ A \ = \ A \ > \ A \ \ G : A \ = | G | . A l s o , we can assume t h a t A < G , and thus A A < G by L e m m a 2.10. T h i s yields
9

G| > \ A A \ =
9

\A\\A A n A

>

G A n A
9

and thus A n A > 1 . Since this holds for a l l g E G, we can apply Zenkov's theorem w i t h B = A to deduce that A n F ( G ) > 1, as wanted.

2 D

63

B y C o r o l l a r y 2.19, we see that if G is n o n t r i v i a l and A is an abelian subgroup of G w i t h \ A \ > \ G : A \ , then A contains a n o n t r i v i a l s u b n o r m a l subgroup of G. I n fact, if A is cyclic, we get more: A contains a n o n t r i v i a l n o r m a l subgroup of G. T h i s is a pretty result of A . L u c c h i n i . (We use L u c c h i n i ' s theorem i n C h a p t e r 3 to prove t h a t every a u t o m o r p h i s m of a finite group G has order less t h a n | G | . ) 2.20. T h e o r e m ( L u c c h i n i ) . L e t A be a c y c l i c p r o p e r s u b g r o u p of a f i n i t e g r o u p G, a n d l e t K = c o r e ( A ) . T h e n \ A : K \ < \ G : A \ , a n d i n p a r t i c u l a r , i f \ A \ > \ G : A \ , t h e n K > 1.
G

P r o o f . W e proceed by i n d u c t i o n on | G | . N o w A / K is a proper cyclic subgroup of G/K, and the core of A / K i n G / K is t r i v i a l . If K > 1, we can a p p l y the inductive hypothesis i n the group G / K to deduce that \ A / K \ < \ ( G / K ) : [ A / K ) \ = \ G : A \ , a n d there is n o t h i n g further to prove. W e can assume, therefore, t h a t K = 1 , and we work to show that \A\ < \G : A \ . A s s u m e that \ A \ > \ G : A \ . Since G > A , we k n o w t h a t G is n o n t r i v i a l , and hence A n F ( G ) > 1 by C o r o l l a r y 2.19. I n p a r t i c u l a r , F ( G ) > 1, so we can choose a m i n i m a l n o r m a l subgroup E of G w i t h E C F ( G ) . T h e n J5l ~lZ(F (G)) > 1, and since E is m i n i m a l n o r m a l i n G, we have E C Z ( F ( G ) ) , and i n p a r t i c u l a r , E is abelian. A l s o , and again using the m i n i m a l i t y of E , it follows that E is an elementary abelian p-group for some p r i m e p . (In other words, x = 1 for a l l x e E . )
p

Since E C Z ( F ( G ) ) , we see t h a t E normalizes the n o n t r i v i a l group A n F ( G ) , and of course A normalizes this subgroup too. T h e n A n F ( G ) < A E ,

64

2.

Subnormality

and since c o r e ( , 4 ) = 1, it follows t h a t A E < G, as is indicated i n the diagram.

G

Write G = G / E .

L e t M = c o r e ^ ) , w i t h M D E , a n d note that M < G G since A E < G, and so by the inductive a n d we have \ A E : M | < \ G : A E \ .

and A M = A E _ A l s o ^ Z _ < hypothesis, \ A : M \ < \ G : A \ , Let

B = A n M , so t h a t 73 is cyclic. W e have \AE :A \ = \ A M :A \ = \ M :A n M \ = \ M : B\,

and hence \ A E : M \ = \ A : B \ . W e conclude t h a t

Suppose that M is abelian, and let ip : M - M be the h o m o m o r p h i s m defined by ^ ( m ) = m . T h e n i ? C ker(v>), and since M = E B by D e d e k i n d ' s l e m m a , it follows t h a t <p(M) = <p{B) C B C A . N o w M < G, and hence <p(M) < G, a n d we conclude that i p ( M ) = 1 since c o r e ( , 4 ) = 1. T h e n f ( B ) = 1, and since B is cyclic, it follows t h a t \ B \ < p . T h e n \ M : B \ < \ B \ < p , a n d since M / B is a p - g r o u p , we deduce that M = B . B u t M < G and M C A , a n d thus M = 1, a n d i n particular, E = 1, w h i c h is a contradiction.
p G

It follows t h a t M is nonabelian, a n d since M / E is cyclic, we conclude that E is not central i n M , and so E n Z ( M ) < P . W e conclude by the m i n i m a l i t y of E t h a t E n Z ( M ) = 1, a n d thus Z ( M ) is cyclic. Since ? is an abelian subgroup of M a n d |M : B \ < \ B \ , it follows by C o r o l l a r y 2.19 t h a t B n F ( M ) > 1. N o w F ( M ) C F ( G ) , and so centralizes F ( M ) , a n d thus B n F ( M ) is a n o n t r i v i a l central subgroup of B E = M . Since Z ( M ) is cyclic, we see t h a t 7 3 n F ( M ) is characteristic i n Z ( M ) < G . T h u s B n F ( M ) is a n o n t r i v i a l n o r m a l subgroup of G contained i n A , a n d this is our final contradiction.

Problems

2D
A

2 D . 1 . L e t G = N A , where N < G , C ( N ) (a) If F(JV) = 1, show t h a t \ A \ < \ N \ .

1 and A is abelian.

(b) If \ N \ a n d \ A \ are coprime, show t h a t \ A \ < \ N \ . 2 D . 2 . L e t G = N A , where N < G, C { N )

A

= 1 a n d A n N = 1. If TV is

n o n t r i v i a l a n d ,4 is cyclic, show that \ A \ < \ N \ .

Chapter 3

Split Extensions

3A G i v e n TV < G , a subgroup H C G is a c o m p l e m e n t for TV i n G i f / V P = G and N n P = 1, and i f /V has a complement i n G , we say that G splits over N . N o t e t h a t i f P complements N i n G , then H = H / { N n P ) = 7 V P / / V = G / / V , and thus a l l complements for N i n G are i s o m o r p h i c to G / i V , and hence they are isomorphic to each other. A n easy example where a group fails to split over some n o r m a l subgroup is G = G , the cyclic group of order 4, where N is the unique subgroup of order 2. It is clear t h a t N is not complemented i n G because if a complement existed, it w o u l d be isomorphic to G/N, w h i c h has order 2. B u t N is the only subgroup of order 2 i n G , and of course, N is not a complement for itself.
4

If P complements N i n G , then each element g G is of the form g = n h , w i t h n e N a n d h e H , and this factorization is unique. (The uniqueness follows v i a an elementary c o m p u t a t i o n , b u t i f G is finite, a n easier argument is to observe t h a t \ G \ = \ N \ \ H \ and so the surjective m a p from pairs (n, h ) to elements n h G G must also be injective.) It is also true, and for similar reasons, t h a t each element of G is uniquely of the form h n , w i t h h G H and n e J V . (To be specific, suppose that g = n h , and we w i s h to write g = h ' n ' , w i t h ri e N and h ' G P . W e have g e N h = h N since J V ^ G , and thus we must have /' = h . T h e equation / i n ' = n h then yields n ' = n . )
h

If P complements N i n G , then clearly, every conjugate of H i n G also complements / V . In general, however, a n o r m a l subgroup N of G m a y have nonconjugate complements, although as we have seen, a l l complements are isomorphic. A n example where complements are not conjugate occurs is i n the K l e i n group G = G x G . If TV is any one of the three subgroups of
2 2

65

66

3. S p l i t

Extensions

order 2, then each of the other two subgroups of order 2 complements N , b u t these complements are not conjugate i n G since G is abelian. Our goal i n this section is the following: given groups N and H , construct a l l (up to isomorphism) groups G having a n o r m a l subgroup N complemented by a subgroup H , where N = N and H = H . (We refer to such a group as a split extension of H by N , but unfortunately, the literature is inconsistent on this point, and this group is sometimes also referred to as a split extension of N by H . ) One such split extension, of course, is the (external) direct product G = N x H , which is the set of ordered pairs (n, h ) w i t h n N and h H , and where m u l t i p l i c a t i o n is componentwise. Here, we take N = {(n, 1) | n N } and H = { ( l , h ) \ h e H } , and it is clear that N = N , HQ = H , N o < G and H is a complement for N i n G. In this case, the complement H also happens to be n o r m a l , and i n general, it is easy to see that i f the n o r m a l subgroup 7V has a n o r m a l complement H i n G, then G is the (internal) direct product of i V and H , and so G = N x H . O u r task, therefore, is to generalize the direct product construction sufficiently so t h a t we get complements H for i V i n G, where H is not necessarily n o r m a l i n G. T h e desired construction is called the "semidirect p r o d u c t " .
0 0 0 0 0 0 0 0 0 Q 0 0 0 0 0 0 0

Before we proceed, however, we digress briefly to discuss extensions that are not necessarily split. G i v e n groups N and H , a group G is said to be an extension of H by i V i f there exists N < G such that N = N and G/N = H . N o t e t h a t i n our definition, the group preceded by the word "of" corresponds to the factor group, and the group preceded by "by" corresponds to the n o r m a l subgroup. A s we mentioned, however, this use of prepositions is sometimes reversed i n the literature, so readers should attempt to determine the precise meaning from the context.
0 0 0

T h e r e is a more fundamental ambiguity here. If G is an extension of H by N , then the n o r m a l subgroup N of G such that N = N and G/N = H is not, i n general, uniquely determined. I n other words, G can be an extension of H by N i n more t h a n one way. In fact, it can happen t h a t w i t h one choice of the n o r m a l subgroup N the extension is split, and w i t h another it is not. P r o p e r l y speaking, therefore, we really should say something like " G is an extension of H by N w i t h respect to the n o r m a l subgroup 7V ", and then it w o u l d make sense to ask i f the extension is split.
0 0 0 0 0

Imagine that the "extension problem" could be completely solved. Suppose, i n other words, that we knew how to construct (up to isomorphism) a l l extensions of i f by A for given finite groups H and N . (We stress t h a t this seems to be far from a realistic assumption.) T h e n , given that we know a l l simple groups, we could recursively construct a l l finite groups. T o do this, assume that we have already constructed all groups of order less t h a n n , where n > 1. E a c h group G of order n has some m a x i m a l n o r m a l subgroup,

3A

67

say N . Since the factor group G / N is simple, every group of order n can be constructed as an extension of some (known) simple group S by some (already constructed) group N of order n/\S\. B u t even if we could do this, we w o u l d still be faced w i t h the very difficult p r o b l e m of deciding whether or not two groups of order n constructed i n this way are isomorphic. R e t u r n i n g to split extensions now, suppose t h a t G is split over a n o r m a l subgroup N , and let H be a complement. Observe t h a t H acts on N v i a conj u g a t i o n i n G, and i n fact, conjugation by h H induces an a u t o m o r p h i s m of N . U H < G, then since i V n H = 1, we k n o w t h a t H centralizes N , and so this conjugation action is t r i v i a l . If we w i s h to construct split extensions t h a t are not merely direct products, therefore, we must consider n o n t r i v i a l conjugation actions of H o n N . Before proceeding w i t h our construction, however, we show that no further i n f o r m a t i o n is needed; the subgroups N and H and the conjugation action of H on N completely determine G up to isomorphism. To state the following uniqueness theorem, we introduce some convenient, but nonstandard notation. Consider isomorphic groups N and 7V where some specific isomorphism is given. F o r n N , we w i l l write n e iVo to denote the image of n under the given i s o m o r p h i s m , so t h a t we can t h i n k of ( ) as the name of the isomorphism. W e w i l l use the same name ( ) to denote a given i s o m o r p h i s m from H to H .
0) 0 0 0 0

3.1. L e m m a . L e t G a n d G be g r o u p s , a n d suppose that N < G is complem e n t e d by H , a n d N < G i s c o m p l e m e n t e d by H . Assume that N = N and H H , w h e r e each of these i s o m o r p h i s m s i s d e n o t e d by ( ) , a n d suppose that
0 0 0 0 0 0 0

(n ) f o r a l l elements from G toG

= (no)'

10

n N a n d h <E H . T h e n t h e r e i s a u n i q u e i s o m o r p h i s m t h a t extends t h e g i v e n i s o m o r p h i s m s N -> JV a n d H - > H .

0 0

P r o o f . E v e r y element g G is uniquely of the form h n , where h H and n N , and so if we seek an i s o m o r p h i s m 9 : G G t h a t extends the given isomorphisms N - > N and H -> H , we have no choice but to define 9 ( g ) = 9 ( h n ) = 9 { h ) 9 { n ) = h n . Since the d e c o m p o s i t i o n g = h n is unique, 9 is well defined, and since every element g G has the form h n w i t h h H o and n N , it is clear that 9 is surjective. It is injective because the decomposition g = h n is unique and h and n are the unique elements of H and N that m a p to h o and n respectively.
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

It remains to show that 0 is a h o m o m o r p h i s m , and so we must compute 9 ( ( h n ) ( k m ) ) , where h , k H and n , m N . W e have (hn){km) = h k k - n k m
l

(hk){n m),

68

3. S p l i t

Extensions

and so by definition, a p p l i c a t i o n of 9 to this element yields {hk)o(n m)

k 0

= h ko{n )om
0

= h k {n ) m
Q 0 Q

= h n komo ,
0 0

and this is exactly 9 { h n ) 9 { k m ) , as required.

In order to construct a l l possible split extensions, we need to examine a little more abstractly the conjugation action of a group H o n a group N . R e c a l l t h a t i f i f acts on a set ft, then each element hH induces a m a p o~}i '. ft > ft, defined by a ^ a - h . Furthermore, a is a p e r m u t a t i o n of ft, and the m a p h ^ a is a h o m o m o r p h i s m from i f into the s y m m e t r i c group Sym(ft). It should be clear that the given action of i f on ft is completely determined by this h o m o m o r p h i s m from i f into Sym(ft).
h h

Suppose now that the set ft of the previous paragraph happens to be a group, w h i c h we now call N . I n this case, it is n a t u r a l to require that for all elements h e H , the m a p a : i V -> N is actually an a u t o m o r p h i s m (and not just a permutation) of N . O f course, since this m a p is a u t o m a t i c a l l y b o t h injective and surjective, it suffices to check t h a t it is a h o m o m o r p h i s m , or e q u i v a l e n t s , that { x y ) - h = { x - h ) { y - h ) for a l l x,y N .
h

G i v e n groups i f and N , we say t h a t i f acts v i a a u t o m o r p h i s m s on N if i f acts on N as a set, and i n a d d i t i o n , ( x y ) - h = { x - h ) ( y - h ) for a l l x,y N a n d h e H . Just as an action of i f on ft determines and is determined by a h o m o m o r p h i s m i f -> Sym(ft), so too, an action v i a automorphisms of i f on N determines and is determined by a h o m o m o r p h i s m i f -> A u t ( i V ) . Perhaps the most n a t u r a l examples of actions v i a automorphisms are where i f is a c t u a l l y a subgroup of A u t ( i V ) and n-h is s i m p l y ( n ) h . Other examples of actions v i a automorphisms occur when b o t h i f and N are subgroups of some group G w i t h i f C N ( J V ) , and where the action is given by conjugation w i t h i n G. In particular, the action of an a r b i t r a r y group G on itself by conjugation is an action v i a automorphisms.
G

It is c o m m o n to use exponential notation instead of dots for actions v i a automorphisms, and so we write n instead of n - h , where n G N and h i f . T h i s is n a t u r a l for conjugation actions, of course, but it is p o t e n t i a l l y confusing for actions v i a automorphisms that are not defined by conjugation inside some given group. T h e following theorem, however, asserts that every action v i a automorphisms is essentially a conjugation action w i t h i n a n appropriate group. T o some extent, therefore, this justifies the exponential n o t a t i o n for actions v i a automorphisms.
h

In order to state our result, we again use the n o t a t i o n ( ) to describe isomorphisms N N and i f -> i f . Technically, this is ambiguous, however, because we w i l l not assume that N and i f are disjoint, and so i f x lies in b o t h of these groups, it may not be clear w h i c h i s o m o r p h i s m to apply
0 0 0

3A

69

i n order to compute x . Nevertheless, we trust t h a t this a m b i g u i t y w i l l not cause confusion.

0

3.2. T h e o r e m . automorphisms. N
0

L e t H a n d N be g r o u p s , a n d suppose T h e n t h e r e exists by a s u b g r o u p H
(n )o =
h

t h a t H acts

on N via subgroup

a group G containing a normal

0

= N , complemented

* H , a n d such
h o

that f o r a l ln N

and h < E H , w e have

(no) .

Here, n (no)
h o

G N i s t h e r e s u l t of l e t t i n g h a c t o n n i n t h e g i v e n a c t i o n , w h i l e
0

i s t h e r e s u l t of c o n j u g a t i n g n

by h

i n G.

It follows by L e m m a 3.1 t h a t the group G of T h e o r e m 3.2 is uniquely determined by N and H and the given action v i a automorphisms of H on TV; it is called the semidirect p r o d u c t of N by H w i t h respect to the given action. A l s o , by L e m m a 3.1, every group G w i t h a n o r m a l subgroup N and complement H is isomorphic to a semidirect p r o d u c t of N by H , and so once we prove T h e o r e m 3.2, it w i l l be fair to say t h a t we have constructed all possible split extensions (up to i s o m o r p h i s m ) . A c o m m o n n o t a t i o n for a semidirect product G of N by H is G = N x > H . ( T h i s n o t a t i o n is incomplete, of course, because it fails to m e n t i o n the specific a c t i o n v i a automorphisms t h a t is essential i n the definition. W h e n we construct a group G = N x < H , we should always specify the action.) N o t e t h a t the semidirect product is a direct product if a n d o n l y if H < G, a n d this happens precisely when the conjugation action of H on JV is t r i v i a l , or equivalent^, w h e n the original action of H on AT is t r i v i a l .
0 0 0

We offer a m n e m o n i c for the correct use of the s y m b o l ' V , w h i c h looks like " x " w i t h an e x t r a little vertical line on the right. If G = N x > H , the little line reminds us t h a t from the point of view of H , the semidirect product can be different from a direct product since H m a y not be n o r m a l . B u t N is n o r m a l i n G, and so from the point of view of N , the semidirect product resembles a direct product. We often identify N w i t h N and H w i t h H v i a the isomorphisms x x , so that the semidirect product G N x > H can be viewed as a split extension w i t h N < G, N H = G and N n H = 1, and where the original action v i a automorphisms of H on N is just the conjugation action w i t h i n G. We stress, however, t h a t that this identification is not always a good idea, and it can lead to confusion. Consider, for example, an a r b i t r a r y group G and let T = G x G w i t h respect to the n a t u r a l conjugation action of G on itself. (Somewhat remarkably, it turns out t h a t T = G x G for a l l groups G.)
0 0 Q

It is seldom necessary or appropriate to refer to the proof of T h e o r e m 3.2. In other words, the specific construction of the semidirect product is largely

70

3. S p l i t

Extensions

irrelevant. T h e point here is t h a t given the three ingredients: a group N , a group H and an action v i a automorphisms of H on N , there is a group G, w h i c h is unique up to i s o m o r p h i s m and satisfies the conclusions of T h e o r e m 3.2. It is usually unnecessary to k n o w exactly how G is constructed. (Indeed, there is more t h a n one way to prove T h e o r e m 3.2, w h i c h means that there is more t h a n one possible construction for the semidirect product.) In order to emphasize this point, we give a few applications of semidirect products before we present the construction. R e c a l l t h a t i n C h a p t e r 2, we defined the dihedral group D = D as a group h a v i n g a n o n t r i v i a l cyclic subgroup C of order n such that a l l elements of D - C are involutions. W e argued using geometric reasoning that such a group a c t u a l l y exists when n > 3. U s i n g semidirect products, we can give a cleaner construction of the dihedral group D and some related groups.
2 n 2 n

Let A be an a r b i t r a r y abelian group, and let T = ( t ) be cyclic of order 2. Since t is the o n l y nonidentity element of T , we can define a n action of T on A by setting x = x ~ for a l l x e A . (Since t = 1 and ( a ; " ) " = x , this is clearly a genuine action.) Because A is abelian, we have ( x y ) ~ = x ~ y ~ , and so our action of T on A is a n action v i a automorphisms. N o w let G = A x T w i t h respect to this action, and identify (as we may) A and T w i t h the corresponding subgroups of the semidirect product G. T h e n A < G and A is complemented by T i n G, and thus \ G : A \ = \T\ = 2, and we have \ G \ = 2 \ A \ . A l s o , the equation x* = x ' , can be viewed as a statement about conjugation i n the semidirect product G. N o w every element g of G - A lies i n the coset A t , and so g = a t for some element a G A . In particular, = a t a t = aa* = a a ' = 1, and so a l l elements of G - A are involutions. In other words, G is a generalized dihedral group and i n particular, i f A is cyclic of order n , then G is dihedral of order 2n.
t l 2 1 1 l l l 1 2 1 g

We can also use the semidirect product construction to prove things. A t first glance, the following may seem almost obvious, but its proof depends on n o n t r i v i a l facts from C h a p t e r 2. 3.3. finite C o r o l l a r y (Horosevskii). L e t a G A u t ( G ) , w h e r e G i s a n o n t r i v i a l g r o u p . T h e n t h e o r d e r o { a ) of a i s less t h a n \ G \ .

P r o o f . L e t A = ( a ) , so t h a t A is a cyclic subgroup of A u t ( G ) and \ A \ = o ( a ) . Since A C A u t ( G ) , there is a n a t u r a l action of A on G by automorphisms, and we construct the semidirect product Y = G x > A w i t h respect to this action. A s usual, we identify G and A w i t h subgroups of T, so t h a t G < T and A is a complement for G. A l s o , the conjugation action of A on G i n T is the original action. Furthermore, since by definition, nonidentity automorphisms of G act nontrivially, it follows that no nonidentity element of A acts t r i v i a l l y on G by conjugation i n T. I n other words, A C \ C { G ) = 1.
V

3A

71

Since G is n o n t r i v i a l and A is cyclic, L u c c h i n i ' s theorem (2.20) applies i n T, a n d we deduce t h a t \ A : K \ < \T : A \ , where K = c o r e ( A ) . B u t K n G C A n G = 1, a n d b o t h AT a n d G are n o r m a l i n T, a n d thus K C A n C ( G ) = 1. T h u s o(<r) = \ A \ = \ A : K \ < \ T : A \ = | G | , as required.
r r

Somewhat similarly, we have the following. 3 .4. C o r o l l a r y . L e t P be a n a b e l i a n p - s u b g r o u p of A u t ( G ) , w h e r e G i s a finite g r o u p of o r d e r n o t d i v i s i b l e by t h e p r i m e p . T h e n P has a r e g u l a r o r b i t o n G. I n p a r t i c u l a r , if G i s n o n t r i v i a l , t h e n \ P \ < | G | . R e c a l l t h a t if a group P acts on a set f l , t h e n a P - o r b i t A i n f l is regular if | A | = | P | , or e q u i v a l e n t s (by the fundamental c o u n t i n g principle) the stabilizer i n P of a n element of A is t r i v i a l . A l s o , i n the s i t u a t i o n of the corollary, if P > 1 and A is a regular P - o r b i t , t h e n A does not contain the identity of G , a n d so | G | > 1 + | A | > | P | . T h i s shows t h a t the last sentence of C o r o l l a r y 3.4 w i l l follow once the existence of a regular P - o r b i t is established. P r o o f of C o r o l l a r y 3 .4. Since P C A u t ( G ) , there is a n a t u r a l action of P o n G v i a automorphisms. L e t T = G x P w i t h respect to this action, and as usual, identify P and G w i t h subgroups of T. T h e n |T : P | = | G | is not divisible by p , a n d so P S y l ( r ) . A l s o , the conjugation a c t i o n of P o n G i n r is the original action, and since P consists of automorphisms, only the identity i n P acts trivially. It follows t h a t P n C ( G ) = 1.
p r

N o w O ( r ) C P , a n d so O ( r ) D G = 1. A l s o , b o t h G and O ( r ) are n o r m a l i n T, a n d hence they centralize each other, and we have O ( r ) C P n C ( G ) = 1. Since the Sylow subgroup P is abelian, it follows by B r o d k e y ' s theorem (1.37) that O ( T ) is a n intersection of two Sylow p subgroups. R e p l a c i n g t h e m by conjugates i f necessary, we can assume t h a t one of these Sylow subgroups is P , a n d the other is P for some element x T. W e thus have P n P = O ( T ) = 1.
p p p p r p x x p

W e can write x = u g , w i t h u G P a n d g G , a n d thus P = P 3 = P , a n d we have W e c l a i m t h a t the P - o r b i t i n G that contains g is regular, a n d this w i l l complete the proof. It suffices to show that the stabilizer i n P of g is t r i v i a l , but since the a c t i o n of P on G is conjugation i n r , what we want to establish is that C ( g ) = 1. W e have C ( g ) = C ( g ) 9 , a n d hence C ( g ) C P n P = 1, as wanted.
x u P P P 9 P

A comparison of the previous two corollaries suggests t h a t perhaps whenever A i s & cyclic subgroup of A u t ( G ) , there is always a regular A - o r b i t i n G . If t h a t were true, it w o u l d be a strong form of C o r o l l a r y 3.3, but alas, it is false.

72

3. S p l i t

Extensions

F i n a l l y , we are ready to prove T h e o r e m 3.2 by a c t u a l l y constructing the semidirect product T = N x H . A c o m m o n way to do this, by analogy w i t h the construction of the external direct product, is to define an appropriate m u l t i p l i c a t i o n on the set of ordered pairs ( h , n ) w i t h h H a n d n N . O n e tedious, but necessary, step using this construction is to check t h a t the denned m u l t i p l i c a t i o n is associative. W e choose to use a different approach, w h i c h avoids checking associativity. T h i s is accomplished by w o r k i n g i n an appropriate s y m m e t r i c group, where, of course, associativity is automatic. P r o o f of T h e o r e m 3.2. L e t ft be the set of ordered pairs ( k , m ) , where k H a n d m N . It is t r i v i a l to check that actions of N and of H on ft are denned by the following: (k,m)-n = (k,mn) and {k, m ) - h = { k h , m ) ,
h h

where ( k , ra) ft a n d n e N and h H . Here, of course, m is the result of letting h act on m i n the given action v i a automorphisms of H on N . Since N acts o n ft, we have a n a t u r a l h o m o m o r p h i s m n i-> n from N into the s y m m e t r i c group Sym(ft), where n is the p e r m u t a t i o n a ^ a-n for a ft. W r i t e N = { n \ n A/'}, so that N C Sym(ft) is a subgroup and ( )o is a surjective h o m o m o r p h i s m from N to A ^ . If n = 1, t h e n ( k , m ) - n = ( k , m ) for a l l k i f and m TV, and i n particular, (1,1) = (1, l ) - n = (1, n ) . It follows that n = l , and so the h o m o m o r p h i s m ( ) has t r i v i a l kernel. It is thus a n i s o m o r p h i s m from N onto N .
0 0 0 0 0 0 0 0

Similarly, i f h H , write h Sym(ft) to denote the m a p a H * a - h , and let H = { h \ h H } . T h e n i T C Sym(ft) is a subgroup and ( ) is a h o m o m o r p h i s m from H onto 7J - ( A s before, we are using the n o t a t i o n ( ) for two different maps.) If h o = 1, then (1,1) = (1, l ) - h = ( h , l ) , a n d thus h = 1 a n d ( ) is an i s o m o r p h i s m from H onto H .
0 0 0 0 0 0 0 h 0 0

N e x t , we argue t h a t ( n ) = (n ) for a l l n N a n d h H . Since this is a statement about elements of Sym(ft), it suffices to check that b o t h { n ) o and {n ) have the same effect on each element of ft. W e have
0 0 h h 0

(k,m)(n )o Also, (k,m)(n )

0 h o

= {k,m)-n

(k,mn ).

= {k,m){ho)- noho = {{{k,m)-h- )-n)-h

= (O^-V^VH
= =
h 1

( k h - \ m ~ \ ) - h (k,mn ),
h h h

where the last equality holds because ( m ~ n ) = mn . ( T h i s is v a l i d because the action of i f on TV is an action v i a automorphisms.) W e conclude

Problems 3 A

73

that ( n ) = ( n ) as claimed, and i n p a r t i c u l a r , H C N ( 7 V ) . We can now define G = N H , so that G is a subgroup of S y m ( f t ) , a n d N < G. A l l t h a t remains is to show that H is a complement for N i n G. Since by definition, G = N H , it suffices to check t h a t N n H = 1. T o this end, suppose that n = h w i t h n G TV a n d / i G H . T h e n
0 0 0 S y m ( Q ) 0 0 0 0 0 0 0 0 0 0 0 0

h o

(/i, 1) = (1, l ) - h = (1, l ) / i and it follows that / i = 1. T h u s n

= (1, l ) n = (1, l ) - n = ( l , n ) ,
0 0

= h

= l , a n d the proof is complete.

We close this section w i t h a description of the w r e a t h product, w h i c h is useful for constructing examples a n d counterexamples. T h e ingredients here are two gronps G a n d H , and a set ft, acted on by G. L e t B be the set of a l l functions from ft into H , a n d make B into a group b y defining m u l t i p l i c a t i o n pointwise. T h u s i f / , g G B , then f g is defined by the formula ( f g ) ( a ) = f ( a ) g ( a ) for a G ft. It is t r i v i a l to check that B is a group, and in fact, B is really just the external direct p r o d u c t of |ft| copies of H . The a c t i o n of G on ft induces an action v i a automorphisms of G on B . If we view B as a direct product, the a c t i o n of G can be described s i m p l y as a p e r m u t a t i o n of coordinates, but it is useful to be more precise. G i v e n / G B and x G G, the function f on ft is defined so t h a t f ( a - x ) = / ( a ) , or e q u i v a l e n t s , f ( a ) = f { a - x ~ ) . (It is routine to show that this defines an a c t i o n v i a automorphisms, and we o m i t the proof.)
x x x l

Now let W = B x < G w i t h respect to this action. T h e n W is the w r e a t h p r o d u c t of H w i t h G, and B , viewed as a subgroup of W, is called the base g r o u p of the w r e a t h product. It is c o m m o n to w r i t e W = H I G, but of course, this n o t a t i o n for the w r e a t h p r o d u c t is defective i n t h a t it does not m e n t i o n the set ft or the action of G on ft. If G is given as a n abstract group, and no set ft is mentioned, t h e n it is understood that H I G is constructed using the "regular" a c t i o n of G. T h i s means t h a t we take ft = G, w i t h G acting by right m u l t i p l i c a t i o n . Sometimes, this is referred to as the regular w r e a t h p r o d u c t of H b y G. Problems 3 A 3 A . 1 . L e t C be a cyclic group of order n divisible by 8, and let z be the unique i n v o l u t i o n i n C. (a) Show t h a t C has a unique a u t o m o r p h i s m a such t h a t c for every generator c of C, a n d show t h a t a has order 2.
a

c~ z

(b) L e t S = C x ( o - ) , so t h a t |5| = 2 \ C \ . Show t h a t half of the elements o i S - C have order 2 and t h a t the other half have order 4. (c) Show t h a t the elements of order 2 i n S - C form a single conjugacy class of S, and s i m i l a r l y for the elements of order 4.

74

3. S p l i t

Extensions

N o t e . T h e group S is the s e m i d i h e d r a l group S D , although i n the literature, the w o r d "semidihedral" is usually reserved for the case where n is a power of 2, and there seems to be no standard name for other members of this family of groups.
2 n

3 A . 2 . L e t S and C be as i n the previous p r o b l e m , and let B be the subgroup of index 2 i n G . Show t h a t the elements of order 4 i n S - C form a coset of B , a n d let Q be the u n i o n of this coset and B . Show t h a t Q is a subgroup of order n . N o t e . Since a l l elements of Q - B have order 4, the i n v o l u t i o n i n B is the unique i n v o l u t i o n i n Q. T h e group Q is the generalized q u a t e r n i o n group Q . (The phrase "generalized quaternion" is often restricted to the case where the order n is a power of 2. T h e undecorated w o r d "quaternion" is usually reserved for Q . )
n 8

3 A . 3 . L e t p be a prime, and suppose that m is a divisor of p-1 w i t h m > 1. Show t h a t there exists a group G of order pm w i t h a n o r m a l subgroup P of order p , and such that G/P is cyclic and Z ( G ) = 1. 3 A . 4 . L e t q be a power of a prime p . Show that there exists a group G of order q ( q - 1) w i t h a n o r m a l elementary abelian subgroup of order q and such t h a t a l l elements of order p i n G are conjugate. H i n t . L e t F be a field of order q and observe that the m u l t i p l i c a t i v e group of F acts v i a automorphisms on the additive group of F . 3 A . 5 . L e t G be a n a r b i t r a r y finite group. Show that G x G = G x G , where the semidirect p r o d u c t is constructed using the n a t u r a l a c t i o n of G on itself by conjugation. 3 A . 6 . R e c a l l t h a t a group action is faithful if the only group element t h a t fixes a l l points is the identity. L e t P be a p-group acting faithfully v i a automorphisms o n a group G w i t h order not divisible by p . Show that P acts faithfully o n some P - o r b i t A i n G . H i n t . Use T h e o r e m 1.38, the generalized B r o d k e y theorem. 3 A . 7 . L e t a A u t ( G ) , a n d suppose that at most two p r i m e numbers divide o { a ) . Show t h a t ( a ) has a regular orbit on G . 3 A . 8 . L e t C be cyclic of order p q r , where p , q and r are distinct o d d primes. (a) L e t s { p , q , r } . Show t h a t A u t ( G ) contains a unique i n v o l u t i o n a fixing elements of G of order s and inverting elements of p r i m e orders different from s.

3B

75

(b) A p p l y i n g (a) three times, w i t h s = p , s = q a n d s = r, we get three involutions i n A u t ( C ) . Show that these together w i t h the identity form a subgroup K C A u t ( C ) of order 4. (c) L e t G = C x i K w i t h the n a t u r a l action, a n d let a be the inner a u t o m o r p h i s m of G induced by a generator of C. Show that ( a ) has no regular orbit on G. 3 A . 9 . L e t W = H l G be a wreath p r o d u c t constructed w i t h respect to a transitive a c t i o n of G on some set Q. L e t B be the corresponding base group. (a) Show t h a t C ( G )
B

is the set of constant functions from fi into H . beB

(b) N o w assume t h a t W is the regular w r e a t h p r o d u c t , so that i ) = G. Let C C G be a n a r b i t r a r y subgroup. Show that there exists such t h a t C ( 6 ) = C.
G

3 A . 10. G i v e n a finite group H a n d a p r i m e p , show t h a t there exists a group G h a v i n g a n o r m a l abelian p-subgroup A such t h a t G splits over A , where G/A = H a n d A = C { A ) .
G

3B The Schur-Zassenhaus theorem is a powerful result t h a t guarantees t h a t a finite group G always splits over a n o r m a l H a l l subgroup N . (Recall that by definition, N is a H a l l subgroup of G i f | / V | a n d \ G : N \ are relatively prime.) The Schur-Zassenhaus theorem says even more: i f N is a n o r m a l H a l l subgroup of G, t h e n a l l complements for N i n G are conjugate. U n f o r t u nately, however, the proof of the conjugacy part of the Schur-Zassenhaus theorem requires the a d d i t i o n a l hypothesis t h a t either N or G / N is solvable. (We review the definition a n d basic properties of solvable groups later i n this section.) If we are w i l l i n g to a p p l y the very deep F e i t - T h o m p s o n odd-order theorem, however, we see t h a t this solvability requirement is not really a l i m i t a t i o n . T h i s is because the a s s u m p t i o n that N a n d G / N have coprime orders guarantees that at least one of these groups has o d d order, and hence is solvable b y the F e i t - T h o m p s o n theorem. T h e conjugacy part of the Schur-Zassenhaus theorem is therefore true unconditionally, a l t h o u g h no direct or elementary proof of this is k n o w n . F o r the purposes of this book, we w i l l assume solvability w h e n appropriate, a n d we w i l l generally not appeal to the F e i t - T h o m p s o n theorem. O f course, i f H is a complement for a n o r m a l subgroup N i n G, t h e n \ H \ = \ G : N \ . It is useful to observe t h a t i n the s i t u a t i o n of the SchurZassenhaus theorem, where |JV| a n d \ G : N \ are coprime, the converse of this

76

3. S p l i t

Extensions

statement is also true. If H C G and \ H \ = \ G : N \ , t h e n i f is a complement for TV i n G. T o see this, observe t h a t \ H \ = \ G : N \ is coprime to \ N \ , and thus TV n H = 1. F r o m this, we deduce that \ N H \ = \ N \ \ H \ = \ G \ , and hence N H = G, and indeed H is a complement for TV as claimed. A l t h o u g h several proofs of the Schur-Zassenhaus theorem are k n o w n , the key step i n a l l of t h e m is to establish the result i n the case where N is abelian. W h a t are essentially routine tools can then be used to reduce the general s i t u a t i o n to this case. ( O f course, the assumption for the conjugacy part that either N or G / N is solvable is used i n the reduction.) O u r first major goal, therefore, is the following. 3.5. T h e o r e m . L e t N < G, w h e r e G i s a finite g r o u p , a n d assume that N is a b e l i a n a n d t h a t \ N \ a n d \ G: N \ a r e c o p r i m e . T h e n N i s c o m p l e m e n t e d i n G, a n d a l l c o m p l e m e n t s a r e c o n j u g a t e . If our goal were to find a certain n o r m a l subgroup i n G, we might proceed by constructing a n appropriate h o m o m o r p h i s m from G, and then considering its kernel. B u t we need a complement for TV, w h i c h w i l l not generally be n o r m a l . W e w i l l construct it as the "kernel" of a certain k i n d of distorted homomorphism. Let G act v i a automorphisms on N , a n d consider a m a p <p : G -> TV. T h e n ip is a crossed h o m o m o r p h i s m i f <p(xy) = <p(x)*<p(y) for a l l elements x a n d y i n G. ( O f course, the exponent y here refers to the given action v i a automorphisms of G on N . ) If the action of G o n TV is t r i v i a l , w h i c h means t h a t n = n for a l l n G N a n d g G G, then a crossed h o m o m o r p h i s m from G to N is s i m p l y an o r d i n a r y h o m o m o r p h i s m . B u t we w i l l see that there are also interesting examples w h e n the action is n o n t r i v i a l . F o r example, let G be an a r b i t r a r y group a n d suppose TV < G, so that G acts on N by conjugation. N o w fix an element n G TV, and define the map <p : G - > N by setting i p ( x ) = [ x , n ] . (Recall t h a t [ x , n] is the c o m m u t a t o r of x and n , w h i c h , by definition, is the element x ^ n ^ x n = { n ~ ) n . ) Since n G N < G, it is clear that <p(x) = [ x , n ] lies i n N , as claimed. N o w ( f ( x y ) = [ x y , n ] , and an easy c o m p u t a t i o n , w h i c h we omit, establishes the identity [ x y , n ] = [ x , n ] [ y , n ] . T h u s ( p ( x y ) = <p(x)v<p(y), and so <p is a crossed h o m o m o r p h i s m .
l x 9 9

B y analogy w i t h kernels of o r d i n a r y homomorphisms, we define the kernel of a crossed h o m o m o r p h i s m ip : G - > N to be the subset ker^) = { i G | (f(x) = 1}. T h e following establishes a few basic facts about crossed h o m o m o r p h i s m s a n d their kernels. 3.6. L e m m a . L e t G a n d N be g r o u p s , a n d suppose t h a t G acts v i a a u t o m o r p h i s m s o n N . L e t ip : G - > N be a c r o s s e d h o m o m o r p h i s m w i t h k e r n e l K . T h e following then hold.

3B

77

(a) <p(l) = 1. (b) i f i s a s u b g r o u p (c) I f x , y e (d) \ G : K \ P r o o f . W e have <P(I) = < p ( i - i ) = <p(i) V ( i ) = v(i) ,
2

of G. = <p(y) if a n d o n l y if K x = Ky.

G, t h e n <p{x) = \<p(G)\.

and it follows t h a t 1 = < p ( l ) . nonempty. If k K , we have

1

T h i s establishes (a) a n d shows that i f is

i = ^,(1) = ^ ( a r ) = w i k f ' ^ i k - )

ifik' ),
f f

where the last equality holds because <p{k) 1 a n d l = 1 for a l l elements g G G. T h u s fc" G i f , a n d i f is closed under inverses. A l s o , i f x , y G I f , then
1

<p(xy) and

= <p(x)My)

- l

l = 1,

so xy G K , a n d hence i f is a subgroup, p r o v i n g (b). Now i f x , y G G and K x = K y , we can write y = kx w i t h k G K . T h e n p(y) = if(kx) = <f(k) <p(x)
x

= <p{x),

where the last equality holds because <p(k) = 1. Conversely, i f <p(x) p(xy- )
1 1 y

= <^(y), we compute that = ^(y) ~^(y)

y

= ^(x) ~^(y)

= ^ ( y y " ) = p ( l ) = 1,

and so xy' G i T . T h u s x G i f y and i f x = i f y , as wanted. F i n a l l y , since <p is constant on right cosets of i f a n d takes distinct values on distinct right cosets, it follows that \ i p ( G ) \ is e x a c t l y the number of such cosets. T h i s proves (d). Our next task is to construct a n appropriate crossed h o m o m o r p h i s m from a finite group G into a n abelian n o r m a l subgroup N . T o do this, we consider the transversals for N i n G. These are the subsets T of G constructed by choosing exactly one member from each coset of N i n G. (Note that since N is n o r m a l , left cosets a n d right cosets are the same, so we need not specify "left transversal" or "right transversal".) Since right m u l t i p l i c a t i o n by a n element g G G permutes the right cosets of a n a r b i t r a r y subgroup of G, it follows that if T is a transversal for N , then T g is also a transversal for N . W e write T to denote the set of a l l transversals for N i n G. If x , y G G, write x = y i f x and y lie i n the same coset of N . T h u s x = y i f and only if N x = Ny, or equivalently, x N = y N . If S, T G T , the relation s = t for s G S and t G T defines a n a t u r a l bijection between S and

78

3. S p l i t

Extensions

T, a n d we use this bijection to define an element of TV that we call d ( S , T ) . (In some sense, d ( S , T ) measures the difference between S a n d T.) W e set d(S,T) =
s=t

Y[s-H,

where the p r o d u c t runs over a l l s E S a n d i E T such t h a t s = t. O f course, if s = t, t h e n s N = t N , a n d so s ~ H TV, a n d hence each factor i n the p r o d u c t lies i n TV. A l s o , because we are assuming t h a t TV is abelian, the order i n w h i c h the factors appear is irrelevant, a n d thus d(S, T ) is a well defined element of TV. Some elementary properties of the function d are given i n the following. 3.7. L e m m a . L e t TV be a b e l i a n a n d n o r m a l i n a finite g r o u p G, a n d l e t S, T a n d U be t r a n s v e r s a l s f o r TV. Then u s i n g the n o t a t i o n defined above, t h e following hold. (a) d ( S , T ) d ( T , U ) (b) d { S g , T g ) = d(S,U).

= d { S , T ) 9 for all g E G .
G : 7 V

(c) d { S , Sn) = n l

l f o r a l l n E TV.

P r o o f . If s = t a n d t = u for s E S, t E T a n d u E U, t h e n TVs N t = N u , and so s = u. Since ( s " i ) ( r u ) = s~ u a n d the order of factors i n the abelian group TV is irrelevant, (a) follows.
1 1 l

The and
l

Now let g E G. If s = t, t h e n TVs = N t , a n d so T V s = N t g , a n d sg = t g . factor { s g ) ~ { t g ) of the p r o d u c t defining d { S g , T g ) is equal to ( s ^ i ) , assertion (b) follows.
5 l 9

F i n a l l y , i f n E TV, we have sTV = (sn)TV, a n d so s = sn. E a c h factor - { s n ) of the p r o d u c t defining d { S , Sn) equals n , and so (c) follows because | 5 | = \ G : TV|.

We are now ready to prove the case of the Schur-Zassenhaus theorem where TV is a n abelian n o r m a l H a l l subgroup of G. P r o o f of T h e o r e m 3.5. F i x an a r b i t r a r y transversal T for TV i n G, a n d define 9 : G -> TV by setting 9 { g ) = d { T , T g ) . W e argue first that 9 is a crossed h o m o m o r p h i s m w i t h respect to the conjugation a c t i o n of G o n TV. If x,y E G, t h e n using L e m m a 3.7(a,b), we compute t h a t d(T, Txy) = d(T, T y ) d { T y , T x y ) = d { T , T y ) d ( T , T x )
y

, since TV is

and so 9 { x y ) = B ( y ) 9 ( x ) v . T h i s , of course, is equal to 9{x)9{y) abelian, a n d so 9 is a crossed h o m o m o r p h i s m , as claimed.

G : 7 V

For n E TV, we have 9 ( n ) = d(T, T n ) = J I b y L e m m a 3.7(c). Since we are assuming t h a t |TV| a n d \ G : TV| are coprime, however, the m a p x i-> \ \ is a p e r m u t a t i o n of the elements of TV. (One way to see this is to observe
G : N X

3B

79

that this m a p has an inverse: the m a p x i-> x , where the integer k is chosen so t h a t k \ G : 7V| = = 1 m o d \ N \ . ) It follows t h a t 9 maps N onto N , and in particular, 9 ( G ) = N . N o w let H = ker (0), so t h a t i f is a subgroup by L e m m a 3.6(b). A l s o , \ N \ = \ 9 { G ) \ = \ G : H \ by L e m m a 3.6(d), and hence \ N \ \ H \ = \ G \ and \ H \ = \ G : N \ . W e have already observed that since \ N \ and \ G : N \ are coprime, this is sufficient to guarantee that i f is a complement for N , as wanted. Now suppose that K is an a r b i t r a r y complement for N i n G. Since N K = G, we see that every coset of N i n G contains a n element of K , and in fact, each coset of N contains a unique element of K because N C \ K = \. T h u s K is a transversal for N , and we let m = d{K, T ) N . W e saw that the restriction of 9 to N is the m a p x xl l , w h i c h is surjective, a n d so we can find n N such that 9 ( n ) = m, a n d thus 9 ( n ~ ) = TO" .
G : A f 1 1

To complete the proof, we show t h a t K H . Since \K\ = \ G : N \ = \ H \ , i t suffices to show that K C i f , or equivalently, t h a t 9 ( k ) = 1 for a l l keK. (Recall that H = ker (0).) F i r s t , observe t h a t i f G AT, then
n n n

m and

fc

= d(K,T)
k

= d(Kk,Tk)
1 k

= d(K,Tk)
1

= d(K, T)d(T, Tk) = W e have = 0(n fc)m

_ 1

m9(k),

so 1 = { m ) - m 9 { k ) = ( m ) - 9 { k ) m .
n 1 1 n

9{k ) = 9{n- kn) = 9(n- k) 9(n)

(m ) 0(fc)m,
_ 1 f c

where we were able to drop the exponent n i n the t h i r d equality because N is abelian. T h u s 9 { k ) = 1, and we are done.
n

N e x t , we prove the existence part of the Schur-Zassenhaus theorem w i t h out assuming t h a t the n o r m a l H a l l subgroup is abelian. 3.8. T h e o r e m . L e t N < G, w h e r e G i s a finite a r e c o p r i m e . T h e n N i s c o m p l e m e n t e d i n G. group a n d \N\ and \ G: N \

P r o o f . W e proceed by i n d u c t i o n on \ G \ . Suppose t h a t there exists a subgroup K < G such that N K = G. T h e n \ K : K C \ N \ = \ G : N \ is coprime to | N | , and hence it is also coprime to \ K n N \ . Since K n N < K , the inductive hypothesis guarantees that K n N is complemented i n AT, and we choose a complement H . T h e n | i f | = \ K : K n JVj = |G : 7V| and as we have seen, this guarantees that i f is a complement for N i n G. We m a y thus suppose t h a t there is no proper subgroup K of G such t h a t N K = G, and so i n particular, N must be contained i n every m a x i m a l subgroup of G. I n other words, N C $ ( G ) , the F r a t t i n i subgroup, and so

80

3. S p l i t E x t e n s i o n s

is nilpotent. A l s o , we c a n assume that N > 1 or else G is the desired

complement for N i n G. _ N o w let Z = Z ( / V ) , _ a n d ^ b s e r v e t h a t 1 < Z < G. W o r k i n g i n the group G = G / Z , we_see t h a t | G : N \ = \ G : 7V| is coprime to |TV| a n d hence is also coprime t o \ N \ . It follows by the inductive hypothesis t h a t G splits over N , a n d we let K be a complement. T h e n G = N K = N K , and hence G = N K . It follows by the e a r l i e r p a r t of the proof t h a t K cannot be proper, and we have K = G. Because K is a complement for N , however, their intersection must be the t r i v i a l subgroup Z . T h u s Z = N n K = N n G = N ,

and hence TV = Z is abelian. T h e result now follows v i a T h e o r e m 3.5.

A s we said, it is necessary to assume that either N or G / N is solvable in order to to prove the conjugacy part of the Schur-Zassenhaus theorem. Perhaps, therefore, a review of some basic facts about solvable groups w o u l d be appropriate, and so we begin w i t h the definition. A (not necessarily finite) group G is solvable if there exist n o r m a l subgroups N such that
t

1 = N

C N iC N

C C N

G,

where N / N ^ is abelian for 1 < i < r . (In particular, a l l abelian groups are solvable since i f G is abelian, we can take A o = 1 and TVi = G . ) A finite chain of nested n o r m a l subgroups of G , extending from the t r i v i a l subgroup to the whole group, is called a n o r m a l series for G , and thus G is solvable precisely w h e n it has a n o r m a l series w i t h abelian factors. R e c a l l that the c o m m u t a t o r s u b g r o u p (or derived subgroup) of a n a r b i t r a r y group G is the subgroup G' generated by a l l commutators [ x , y ] = x ~ y ~ x y , w i t h x,y G G . Since [ x , y ] = 1 i f and o n l y i f x and y commute, we see that G is abelian i f and o n l y if G' = 1. A l s o , if tp : G -> H is a surjective h o m o m o r p h i s m , then ip maps the set of commutators i n G onto the set of commutators i n H , and thus <p(G') = H ' . It follows that H is abelian if a n d o n l y i f G' C ker(</?), and i n particular, i f N <3 G , t h e n G / / Y is a b e l i a n precisely w h e n G' C AT. I n other words, the derived subgroup of G is the unique smallest n o r m a l subgroup of G having a n abelian factor group.
x x

The subgroup ( G ' ) ' is usually denoted G", and its derived subgroup, is G'". T h i s n o t a t i o n r a p i d l y gets unwieldy, however, and so we write G ^ for the n t h derived subgroup. I n other words, G ^ = G a n d for n > 0, we define G^ = (G^" ))'. T h e subgroups G ^ are characteristic, and they constitute the d e r i v e d series of G .
1

The

connection w i t h solvability is the following.

3B

81

3.9. L e m m a . A g r o u p G i s s o l v a b l e if a n d o n l y i f G ^ = 1 f o r some m > 0. P r o o f . If G is solvable, choose n o r m a l subgroups N i such t h a t 1 = N

0

integer

C C N

= G,
{

where N i / N i - i is abelian for 1 < i < r , a n d thus ( N i ) ' C N - i for subscripts i n this range. It is clear, however, t h a t i f i f C G, then i f ' C G ' , and thus since G ' = ( N ) ' C 7 V _ i , we have (if r > 2) t h a t G " C ( 7 V _ i ) ' C i V _ . C o n t i n u i n g like this, we get G ( ) C / V _ for m < r, a n d i n particular, G W C N o = 1.
r r r r 2 m r m

Conversely, suppose t h a t G ( ) = 1. Since the subgroups G are characteristic i n G , they are certainly n o r m a l , a n d we have the n o r m a l series 1 = G
( m )

c G

c c G

( 0 )

= G.

T h e factors here are abelian, a n d so G is solvable.

T h e characterization of solvability i n terms of the derived series allows us to prove a number of useful facts. 3.10. Lemma. L e t G be a n a r b i t r a r y then C G group.

(a) I f H Q G ,

^ f o ra l l m > 0 . then < p ( G ^ ) =

(b) I f < p : G - > H i s a f o r a l l m > 0.

surjective h o m o m o r p h i s m ,

(c) If G i s s o l v a b l e a n d H C G, t h e n H i s s o l v a b l e . (d) I f G i s s o l v a b l e a n d N < G, t h e n G / N i s s o l v a b l e . (e) L e t N < G a n d suppose solvable. that N a n d G / N are solvable. Then G is

P r o o f . If H C G , then H ' C G' a n d (a) follows b y repeating this observation. Similarly, (b) follows by repeated a p p l i c a t i o n of the fact that <p(G') = H ' i f ip : G - > H is a surjective h o m o m o r p h i s m . N o w suppose t h a t G is solvable, so t h a t G^ = 1 for some integer n . If H C G , then C G ( ) = 1 by (a), and thus H is solvable, p r o v i n g (c). If <p : G -> i f is a surjective h o m o m o r p h i s m , then i f ( ) = ( ^ ( G ^ ) = ^(1) = 1, and H is solvable. In particular, if N < G a n d we take <p to be the canonical h o m o m o r p h i s m of G onto G/N, w e o b t a i n (d).
n n

For (e), let homomorphism. n , a n d thus G ( ) for some integer is solvable.

n

H = G / N a n d again let y : G -> i f be the canonical Since i f is solvable, i p ( G ^ ) = i f (") = 1 for some integer C ker(y>) = TV. A l s o , since / V is solvable, we have 7V"( ) = 1 m, and thus G ( ) = ( G ^ ) ^ C / V ( ) = 1, a n d thus G
m m + n m

82

3. S p l i t

Extensions

R e c a l l t h a t an abelian p - g m u p P is elementary abelian i f x = 1 for all x P . A useful fact about solvable groups is that a m i n i m a l n o r m a l subgroup of a finite solvable group must be an elementary abelian p-group for some p r i m e p . T h e following is a slightly more general version of this. 3.11. L e m m a . L e t M be a s o l v a b l e m i n i m a l n o r m a l s u b g r o u p of a n a r b i t r a r y g r o u p G. T h e n M i s a b e l i a n , a n d if M i s f i n i t e , i t i s a n e l e m e n t a r y a b e l i a n p - g r o u p f o r some p r i m e p . P r o o f . Since M is solvable, its derived series must eventually reach 1, and thus since M > 1, we conclude that M ' < M . B u t M ' is characteristic i n M , a n d hence it is n e r m a l i n G. It follows by the m i n i m a l i t y of M t h a t M ' = 1, a n d thus M is abelian, as wanted. N o w assume t h a t M is finite, a n d let u M be an element of p r i m e order, say p . Since M is abelian, the set S = { x M | x = 1} is a characteristic subgroup of M a n d hence it is n o r m a l i n G. A l s o , S > 1 since u S, and thus S = M by the m i n i m a l i t y of M . T h i s completes the proof.
p

F i n a l l y , we mention that the derived length of a solvable group G is the smallest integer n such t h a t G^ = 1. (Thus, for example, a solvable group has derived length 1 precisely when it is n o n t r i v i a l a n d abelian.) T h e proof of L e m m a 3.10 shows that subgroups and h o m o m o r p h i c images of a solvable group w i t h derived length n have derived lengths at most n . A l s o , if N < G, where N is solvable w i t h derived length m a n d G / N is solvable w i t h derived length n , then the derived length of G is at most m + n . We are now ready to complete our proof of the Schur-Zassenhaus theorem. 3.12. T h e o r e m . L e t N < G, w h e r e G i s finite a n d \ N \ a n d \ G : N \ a r e c o p r i m e , a n d assume either that N is solvable o r that G / N is solvable. Then all complements f o r N i nG are conjugate. P r o o f . W e proceed by i n d u c t i o n on \ G \ . L e t U C G be arbitrary. W e w i l l show t h a t U satisfies the hypotheses of the theorem w i t h respect to its n o r m a l subgroup U n N , a n d also that i f U contains a complement H for N i n G, then H is also a complement for U P i N i n U. It w i l l then follow v i a the inductive hypothesis that i f U is proper i n G a n d contains two complements for N , these complements are conjugate i n U a n d hence i n G, as wanted. N o w \ U n N \ divides \ N \ and \ U : U n N \ = \ U N : N \ divides \ G : N \ , a n d thus \ U C \ N \ a n d \ U : U n N \ are coprime. A l s o , either G / N is solvable, in w h i c h case U / { U n N ) = U N / N C G / N is solvable, or else N is solvable, a n d so U n N is solvable. If U contains a complement H for N i n G, then \ H \ divides |J7| a n d is coprime to \ U n N \ , and hence \ H \ divides \ U : U n N \ . B u t \ U : U n TV| divides \ G : N \ = \ H \ , a n d so we have \ H \ = \U : U n I f |,

3B

83

a n d therefore i f complements U n N i n U. T h i s establishes the assertions of the previous paragraph. Next^we_consider factor groups. Suppose L < G , a n d write G = G / L , so that N < G . Since N is a h o m o m o r p h i c image of N , it is solvable if N is, and its order divides \ N \ . A l s o G / N = G / N L is a h o m o m o r p h i c image of G/N, and_so it is solvable i f G / N is, a n d its order divides \ G : N \ . I n other words, G satisfies the hypotheses w i t h respect to the normal_subgroup N . If_H is a complement for N i n G, t h e n \ H \ _ \ s coprime to \ N \ , and so AT n i f = 1. A l s o , N H = ~NH = G, and thus H is a complement for N i n G. L e t i f and K be complements for N i n G . If L _ > 1, t h e n by the inductive hypothesis a p p l i e d i n G, we deduce t h a t i f a n d K are conjugate i n G , and thus H L and ATL are conjugate i n G. W e can thus w r i t e i f f , = ( K L ) for some element g e G, and hence H L contains b o t h H and K , each of w h i c h is a complement for N i n G. If i f L < G, therefore, it follows by the result of the first paragraph of the proof t h a t i f a n d K are conjugate, and thus i f and K are conjugate, as required. W e c a n thus assume that H L = G for every nonidentity n o r m a l subgroup L of G.
9 9 9

Suppose t h a t N is solvable, and note t h a t we c a n assume t h a t N > 1, or else i f = G = K and there is n o t h i n g to prove. L e t L C N be a m i n i m a l n o r m a l subgroup of G, so that L is abelian b y L e m m a 3.11. B y the result of the previous paragraph, i f L = G, and since L C N a n d \ N \ = \ G : i f |, it follows t h a t N = L , a n d so N is abelian a n d we are done by T h e o r e m 3.5. F i n a l l y , assume that G / N is solvable. W e can assume that N < G, or else i f = 1 = K a n d there is n o t h i n g to prove. L e t M / N be m i n i m a l n o r m a l i n G/N, a n d a p p l y L e m m a 3.11 to conclude t h a t M / N is a p-group for some prime p . A l s o , since p divides \ G : 7V|, we see t h a t p does not divide |7V|. Since N H = G and N C M , D e d e k i n d ' s l e m m a yields M = N ( M n i f ) , a n d it follows that M n i f complements N i n M . I n p a r t i c u l a r | M n i f | = | M : 7V|, a n d so M n i f is a p - g r o u p . A l s o , | M : M n i f | = \ N \ is not divisible by p , a n d we conclude that M n i f is a S y l o w p-subgroup of M . Similarly, M n K is a Sylow p-subgroup of M , and thus by the Sylow C - T h e o r e m , M n i f = ( M n A T ) = M n A T for some element m G M .
m m

N o w write L M n i f = M C \ K , and note t h a t L < i f and L o i f since M < G. T h u s i f and A T are complements for N i n G t h a t are contained i n N ( L ) . If N ( L ) < G , we are done by the first part of the proof. W e can assume, therefore, that L < G , a n d since L > 1, we have L H = G. B u t L C i f , and so i f = G a n d thus also K = G and we are done i n this case too.
m G G

84

3. S p l i t

Extensions

Perhaps it is of interest to note that the conjugacy part of T h e o r e m 3.5 was used i n the proof of T h e o r e m 3.12 only i n the case where N is solvable. Conjugacy i n the case where G / N is solvable u l t i m a t e l y depends on the conjugacy of Sylow subgroups, but not on T h e o r e m 3.5. Problems 3B 3 B . 1 . Show t h a t a m a x i m a l subgroup of a finite solvable group must have prime power index. H i n t . L o o k at a m i n i m a l n o r m a l subgroup and work by i n d u c t i o n . 3 B . 2 . Suppose 1 = N < N < < N = G, where the factors N i / N ^ are abelian for 1 < i < r . Show that G is solvable. (Note that we are not assuming t h a t the subgroups iVj are n o r m a l i n G.)
0 x r

3 B . 3 . L e t G be finite a n d let 1 = 7V < N < < N = G, where the factors ZVi/TVi-i are simple for 1 < i < r . Show that G is solvable i f a n d only i f these factors a l l have p r i m e order.
0 x r

N o t e . R e c a l l t h a t a s u b n o r m a l series such as { N } , where the factors are simple, is a c o m p o s i t i o n series for G, a n d note that if G is finite, such a series necessarily exists. A l t h o u g h G m a y have several different composit i o n series, the J o r d a n - H o l d e r theorem asserts that the factors N i / N ^ are uniquely determined by G (except for the order i n w h i c h they occur). T h i s p r o b l e m says t h a t solvable finite groups are exactly the groups for w h i c h every c o m p o s i t i o n factor has p r i m e order. 3 B . 4 . L e t N < G , where \ N \ a n d \ G : N \ are coprime. L e t U C G, where \ U \ divides \ G : N \ , a n d assume that either N or U is solvable. Show that U is contained i n some complement H for N i n G. 3 B . 5 . Suppose t h a t P e S y l ( G ) and t h a t P C Z ( G ) . Show that the set X of elements of G w i t h order not divisible by p is a subgroup of G, and t h a t G = X xP .
p

3 B . 6 . L e t N < G a n d g G, a n d suppose that w h e n N g is viewed as an element of G/N, it has has order TO. (a) Show t h a t there exists h N g such t h a t every prime divisor of o { h ) divides m. (b) IfTOand \ N \ are coprime, show that the element h of part (a) must have order TO. (c) N o w assume t h a t N g is conjugate to its inverse i n G / N and that TO a n d \ N \ are coprime. L e t h N g be as i n (a). Show t h a t h is conjugate to its inverse i n G.

Problems 3 B

85

H i n t . F o r (a), find a cyclic rr -subgroup C such t h a t N C = N ( g ) , where rr is the set of p r i m e divisors of m. F o r (c), let x G , where ( N g ) = (Ng)-\ Observe t h a t x normalizes N ( h ) . Consider the subgroup ( h ) .
N x x

3 B . 7 . A group G is supersolvable i f there exist n o r m a l subgroups N i w i t h 1 = 7V C TVi C C N

0 r

= G,

a n d where each factor N i / N ^ is cyclic for 1 < i < r. Clearly, supersolvable groups are solvable, and it is routine to check t h a t subgroups a n d factor groups of supersolvable groups are supersolvable. Suppose now t h a t G is finite a n d supersolvable. (a) Show that the order of every m i n i m a l n o r m a l subgroup of G is prime. (b) Show that the index of every m a x i m a l subgroup of G is prime. H i n t . F o r (b), look at a m i n i m a l n o r m a l subgroup and w o r k by i n d u c t i o n . 3 B . 8 . L e t G be finite, and assume t h a t every m a x i m a l subgroup of G has p r i m e index. Show t h a t G is solvable. Show also t h a t G has a n o r m a l Sylow p-subgroup, where p is the largest p r i m e divisor of \ G \ , a n d that G has a n o r m a l g-complement, where q is the smallest p r i m e divisor of \ G \ . N o t e . I n fact, G must be supersolvable, b u t this theorem of B . H u p p e r t is m u c h harder to prove. R e c a l l t h a t a g -complement i n a group G is subgroup w i t h g-power index a n d h a v i n g order not divisible by q. In other words, it is a subgroup whose index is the order of a Sylow g-subgroup. 3 B . 9 . L e t G be finite a n d supersolvable. If n is a divisor of \ G \ , show t h a t G has a subgroup of order n . H i n t . L o o k at a m i n i m a l n o r m a l subgroup a n d work by i n d u c t i o n . 3 B . 1 0 . L e t G be finite and supersolvable, a n d suppose t h a t p is the largest p r i m e divisor of \ G \ . Show that G has a n o r m a l S y l o w p-subgroup. H i n t . L o o k at a m i n i m a l n o r m a l subgroup a n d work by i n d u c t i o n . 3 B . 1 1 . A s usual, let $ ( G ) be the F r a t t i n i subgroup of the finite group G. Show t h a t every p r i m e divisor of | $ ( G ) | also divides | G : $ ( G ) | . 3 B . 1 2 . L e t G be solvable, and assume t h a t $ ( G ) = 1. L e t M be a m a x i m a l subgroup of G , a n d suppose that H C M . Show t h a t G has a subgroup w i t h index equal to | M : H \ . 3 B . 1 3 . Show t h a t every finite group contains a unique largest solvable normal subgroup.

86

3. S p l i t

Extensions

3 B . 1 4 . L e t F F ( G ) , where G is finite, and let C = C ( F ) . Z ( F ) is the unique largest solvable n o r m a l subgroup of C.

G

Show that

H i n t . P r o b l e m I D . 19 is relevant. N o t e . If G is solvable, it follows that C = Z ( F ) . solvable groups, F ( G ) 5 C ( F ( G ) ) .

G

T h i s shows that for

3 B . 1 5 . (Berkovich) L e t G be solvable, and let i f < G be a proper subgroup h a v i n g the smallest possible index i n G . Show that H < G . 3C T h e r e exists a n extensive and well developed theory of finite solvable groups to w h i c h we give only superficial mention here. (Readers interested i n pursuing this might w i s h to consult the comprehensive b o o k on solvable groups by D o e r k a n d Hawkes.) U n d e r l y i n g this theory are some results of P h i l i p H a l l analogous to the Sylow E - , C - and D-theorems, but where sets of primes replace i n d i v i d u a l primes. H a l l ' s results are not especially difficult to prove directly, but they also follow easily from the Schur-Zassenhaus theorem, and that is w h y we present t h e m here. R e c a l l that i f vr is a set of primes, then a vr-group is a finite group such t h a t a l l primes d i v i d i n g its order lie i n TT , and a vr-subgroup of a group G is s i m p l y a subgroup t h a t happens to be a vr-group. A H a l l vr -subgroup of a finite group G is a rr-subgroup w i t h index involving no p r i m e of vr. For convenience i n discussing H a l l ' s theorems a n d related topics, we say t h a t an integer m is a vr-number if every prime divisor of m lies i n vr. A l s o , we w r i t e vr' to denote the complement of vr i n the set of a l l p r i m e numbers, so that a vr'-number is an integer w i t h no divisors i n vr. (In the case where TT consists of a single p r i m e p , we generally w r i t e p ' i n place of { p } ' . ) A subgroup i f of G , therefore, is a H a l l rr-subgroup of G precisely w h e n \ H \ is a vr-number and | G : i f | is a vr'-number, or equivalently, \ H \ is the largest vr-number d i v i d i n g | G | . O f course, if vr consists of just one prime, so that vr = { p } , then a H a l l vrsubgroup is just a Sylow p-subgroup, a n d we k n o w by the Sylow E-theorem that such a subgroup a c t u a l l y exists for every finite group. B u t if |vr| > 1, then an a r b i t r a r y finite group G can fail to have a H a l l vr-subgroup. T h e s i t u a t i o n is different, however, if G is solvable. 3.13. T h e o r e m ( H a l l - E ) . Suppose t h a t G i s a finite s o l v a b l e g r o u p , a n d l e t TT be a n a r b i t r a r y set of p r i m e s . T h e n G has a H a l l t r - s u b g r o u p . P r o o f . W e c a n certainly assume that G > 1, and we proceed by i n d u c t i o n on | G | . L e t M be a m i n i m a l n o r m a l subgroup of G a n d let J f be a H a l l vrsubgroup of G = G / M , where H D M . ( O f course, i f exists by the inductive

3C

87

hypothesis.) T h e n \ H : M \ = \ H \ is a Tr-number a n d | G : H \ = \ G : H \ is a vr'-number. B y T h e o r e m 3.11, we know that M is a p-group for some p r i m e p . If p G vr, then since \ H \ = \ M \ \ H : M | , it follows t h a t \ H \ is a vr-number, and thus H is a H a l l vr-subgroup of G, as wanted. If p & TT , on the other h a n d , t h e n | M | and \ H : M \ are coprime, and hence by the Schur-Zassenhaus theorem, M has a complement K i n H . T h e n \ K \ = \ H : M \ is a vr-number and | G : AT| = | G : H \ \ H : K \ = \ G : H \ \ M \ is a rr '-number. I n this case, K is the desired H a l l vr-subgroup of G . W e used the Schur-Zassenhaus theorem to prove T h e o r e m 3.13, but n o t h i n g t h a t deep is really needed. In fact, the H a l l E - t h e o r e m can be proved using a fairly elementary argument. N e x t , we present ( w i t h a somewhat sketchy proof) the H a l l C-theorem, but we defer the H a l l D - t h e o r e m to the problems at the end of this section. 3.14. T h e o r e m ( H a l l - C ) . Suppose t h a t G i s a finite s o l v a b l e g r o u p , a n d l e t TT be a n a r b i t r a r y set of p r i m e s . Then a l l H a l l T x - s u b g r o u p s of G a r e conjugate. P r o o f . A s i n the previous proof, we proceed b y i n d u c t i o n on | G | . A s s u m i n g t h a t G > 1, choose a m i n i m a l n o r m a l subgroup M of G , w h i c h must be a p-group for some p r i m e p . L e t H a n d K be H a l l vr-subgroups of G , and observe that H a n d K _ a v e Ha]l_ vr-subgroups of G = G / M . B y the i n d u c t i v e hypothesis, therefore, H and K are conjugate i n G , and it follows that H M a n d K M are conjugate i n G . W e can thus w r i t e H M = { K M ) for some element g G G .
9

If p G vr, then \ H M \ and \ K M \ are vr-numbers. B u t \ H \ = \ K \ is the largest vr-number d i v i d i n g | G | , and it follows t h a t H = H M and K = K M , a n d so H = K , as wanted. If p vr, then H a n d K are complements i n H M for the n o r m a l H a l l subgroup M . T h e conjugacy part of the SchurZassenhaus theorem then yields the result.
9 9

One might wonder how essential solvability is i n the H a l l E - t h e o r e m . C e r t a i n l y , some nonsolvable groups have H a l l vr-subgroups for some interesting sets vr of primes. ( B y "interesting" here, we mean cases where vr contains more t h a n one, but not a l l of the p r i m e divisors of the group order. In other cases, the existence of a H a l l vr-subgroup is either a t r i v i a l i t y , or is guaranteed by Sylow's theorem.) F o r example, the simple group A contains the H a l l {2, 3}-subgroup A , and the simple group of order 168 contains a H a l l {3, 7}-subgroup and a H a l l {2, 3}-subgroup.
5 4

W e say that G satisfies E i f it contains a H a l l vr-subgroup. W e k n o w t h a t solvable groups satisfy E for a l l sets vr of primes, a n d perhaps surprisingly,
T T

88

3. S p l i t

Extensions

the converse is true: if G satisfies E for a l l IT, then G must be solvable. In fact, G is solvable i f it satisfies E for a l l p r i m e sets of the form TT = p ' , where p is an a r b i t r a r y prime. (Recall that p ' is the complement of { p } i n the set of a l l primes.) A H a l l p'-subgroup of G is called a p - c o m p l e m e n t in G, and of course, i f \ G \ is not divisible by p , then G is a p-complement of itself. T h e c o n d i t i o n E > has no content i n this case, and so we can state the converse of the H a l l E - t h e o r e m as follows.
v T p

3.15. T h e o r e m .

Suppose

that a

finite

g r o u p G has a p - c o m p l e m e n t f o r a l l

p r i m e d i v i s o r s p of \ G \ . T h e n G i s s o l v a b l e . W h a t does this converse of H a l l ' s E - t h e o r e m say i n the case where \ G \ involves just two (or fewer) primes? If \ G \ = p q , where p and q are prime a n d a and b are nonnegative integers, then a Sylow g-subgroup of G is a p-complement a n d a Sylow p-subgroup of G is a g-complement. I n this situation, Sylow's theorem guarantees that the hypotheses of T h e o r e m 3.15 are satisfied, and thus the theorem says that every group of order p q is solvable. T h i s is B u r n s i d e ' s famous pY-theorem, w h i c h we are certainly not prepared to prove at this point. (We present a proof i n C h a p t e r 7, however.) B u t i f we are w i l l i n g to assume B u r n s i d e ' s theorem, it is not h a r d to establish T h e o r e m 3.15 i n its full generality.
a b a b

We need a few p r e l i m i n a r y results. T h e first of these is fairly standard, and it also appears i n the gronp-theory review i n the appendix. 3.16. L e m m a . L e t H , K C G, w h e r e G i s a \G : H \ and \G : K \ are coprime. Then G = H K finite g r o u p , a n d suppose that

a n d \ H : H n K \ = \ G: K \ .

P r o o f . L e t D = H n K , a n d observe that since \ G : D \ = \ G : H \ \ H : D \ , the index \ G : D \ is divisible by \ G : H | . Similarly, \ G : D \ is divisible by \ G : K \ , and since we are assuming t h a t \ G : H \ and \ G : K \ are coprime, we conclude t h a t their product divides \ G : D \ . In particular, \ G : H \ \ G : K \ < \ G : D \ , and hence \ G \ < \ H \ \ K \ / \ D \ = \ H K \ . It follows t h a t H K = G, as wanted. A l s o , since equality holds here, we o b t a i n \ G : K \ = \ G \ / \ K \ = \ H \ / \ D \ = \ H : H n K \ , as wanted. We m e n t i o n t h a t this proof of L e m m a 3.16 is v a l i d o n l y i f \ G \ is finite, but i n fact, the l e m m a is true even if G is infinite, assuming, of course, t h a t H and K have finite coprime indices. The following is due to H . W i e l a n d t . suppose If each

3.17. T h e o r e m . L e t H , K , L C G, w h e r e G i s a finite g r o u p , a n d that the indices \ G : H \ , \ G : K \ a n d \ G: L \ are p a i r w i s e coprime. o f H , K and L is solvable, then G is solvable.

3C

89

P r o o f . Since the result is t r i v i a l i f | G | = 1, we c a n proceed by i n d u c t i o n on \ G \ . Suppose t h a t N < G, a n d consider the subgroups H , K and L of G = G/N. Since these are h o m o m o r p h i c images of the solvable groups H , K and L , respectively, each of t h e m is s o l v a b l e ^ A l s c J G : H | = | G : N ~ H \ = | G : N H \ , w h i c h divides | G : H \ . S i m i l a r l y , | G _ ^ / ^ d i v i d e s | G : K \ and | G : L \ divides | G : L | , a n d ^ o the indices of H , K a n d L are pairwise coprime. If N > 1, therefore, G is solvable b y the i n d u c t i v e hypothesis. If H = 1, then | G | = | G : i f | is relatively p r i m e to | G : K \ , and thus G = K is solvable. W e can assume, therefore, t h a t H > 1, and we choose a m i n i m a l n o r m a l subgroup M of H . B y L e m m a 3.11, we k n o w that M is a p-group for some p r i m e p, and since p cannot d i v i d e b o t h | G : Af| and | G : L | , we can assume that p does not d i v i d e | G : AT|. I n other words, K contains a full Sylow p-subgroup of G , and by Sylow theory, some conjugate of t h a t Sylow subgroup contains the p-subgroup M . T h u s M C K for some element g G G .
9

N o w | G : K \ = \ G : K \ is relatively p r i m e to | G : i f | , a n d hence by L e m m a 3.16, we have G = H K . L e t s G be arbitrary, and w r i t e x = w , where u e H a n d v e K . T h e n M * = M = M C K , where the second equality holds because u H and M < H . A l l conjugates of M , therefore, lie i n K , and hence the subgroup M t h a t they generate is contained i n K , w h i c h is solvable. T h e n M is a nonidentity solvable n o r m a l subgroup of G , and also, G / M is solvable by the first paragraph of the proof. It follows b y L e m m a 3.10(e) that G is solvable, as required.
9 9 U 1 ; v 9 9 G 9 G G

P r o o f of T h e o r e m 3.15 (assuming B u r n s i d e ) . W e proceed by induct i o n on the number n of different primes t h a t d i v i d e | G | . If n = 1, then G is a p-group, a n d hence is solvable, a n d i f n = 2, then G is solvable by B u r n s i d e ' s theorem. W e can assume, therefore, t h a t at least three primes p, q and r d i v i d e | G | , and we choose a p - c o m p l e m e n t H , a g-complement K and a n r-complement L i n G . T h e indices of these subgroups are respectively, a power of p, a power of q and a power of r , and so they are pairwise coprime. If we can show t h a t H , K and L are solvable, it w i l l follow by T h e o r e m 3.17 t h a t G is solvable, and we w i l l be done. T h e p r i m e divisors o f \ H \ are just the p r i m e divisors of | G | other t h a n p, and so there are exactly n - 1 of t h e m . L e t s be one of these p r i m e divisors of \ H \ , a n d let Q be an s-complement i n G . T h e n | G : H \ is a power of p and | G : Q\ is a power of s, and so these indices are coprime, a n d it follows by L e m m a 3.16 t h a t \ H : H f ] Q \ = \ G : Q\, w h i c h is the full power of s d i v i d i n g | G | . Since this is also the full power of s d i v i d i n g \ H \ , it follows t h a t H n Q is a n s-complement i n H . T h i s shows t h a t H has a n s-complement for each of the n - 1 primes s d i v i d i n g its order, and so H is solvable by the inductive hypothesis. Similarly, K and L are solvable, a n d the proof is complete.

90

3. S p l i t

Extensions

W e have seen t h a t for every nonsolvable finite group G, there is a prime set rr for w h i c h G fails to have a H a l l rr -subgroup. H o w about conjugacy? Since H a l l rr -subgroups do sometimes exist i n nonsolvable groups, we can ask whether or not (for a fixed prime set rr) a l l H a l l rr -subgroups of a nonsolvable group G must be conjugate. T h e answer is "no"; they need not even be isomorphic. In the simple group PSL(2,11) of order 660 = 2 - 3 - 5 - l l , for example, there exists a H a l l {2, 3}-subgroup isomorphic to the alternating group A , a n d there is another H a l l {2, 3}-subgroup isomorphic to the d i hedral group D . These groups of order 12 are certainly not isomorphic. W e close this discussion by m e n t i o n i n g (without proof) a lovely theorem of W i e l a n d t w h i c h asserts t h a t if G has a n i l p o t e n t H a l l rr -subgroup, t h e n a l l H a l l rr -subgroups of G are conjugate.
2 A 1 2

Problems 3C 3 C . 1 . L e t U C G be a rr -subgroup, where rr is a set of primes a n d G is solvable a n d finite. Show t h a t U is contained i n some H a l l rr -subgroup of G. H i n t . Use P r o b l e m 3B.4. N o t e . T h i s , of course, is the H a l l D-theorem. 3 C . 2 . L e t 7t be a set of primes, a n d suppose that G has a p-complement H for each p r i m e p e r t . Show that C\ H is a H a l l rr '-subgroup of G.
p pen v

3 C . 3 . A Sylow system i n a group G is a set S of Sylow subgroups of G, one chosen from S y l ( G ) for each prime divisor p of a n d such t h a t P Q = Q P for a l l P , Q 5.
p

(a) Show t h a t if G has a Sylow system, t h e n it has a H a l l rr -subgroup for every set rr of primes. (b) If G is solvable, prove t h a t it has a Sylow system. H i n t . F o r (b), choose a p-complement H of G for each p r i m e divisor of \ G \ , a n d consider intersections of a l l but one of these p-complements.
p

3 C . 4 . L e t M be m i n i m a l n o r m a l i n a finite solvable group G, a n d assume that M = C ( M ) .

G

Show that G splits over M a n d that a l l complements

for M i n G are conjugate. H i n t . L e t L / M be m i n i m a l n o r m a l i n G / M , a n d note t h a t L / M is a qgroup for some p r i m e q. Show that q does not divide | M | , a n d consider a Sylow q subgroup of L .

3D

91

3 C . 5 . L e t i f be a m a x i m a l subgroup of G , where G is solvable, a n d assume t h a t c o r e ( i f ) = 1. Show that G has a unique m i n i m a l n o r m a l subgroup M , t h a t H complements M a n d t h a t M = C { M ) . Deduce t h a t i f K is also m a x i m a l i n G w i t h t r i v i a l core, t h e n H a n d K are conjugate i n G.
G G

3 C . 6 . L e t G be finite and solvable, a n d suppose t h a t x , y , z G G have orders t h a t are pairwise relatively prime. If xyz = 1, show t h a t x = 1, y = 1 and z = l. H i n t . W o r k by i n d u c t i o n o n the derived length of G. 3 C . 7 . A C a r t e r s u b g r o u p of a solvable group G is a nilpotent subgroup Csuch thatC = N (C).
G

(a) Show t h a t every solvable group has a C a r t e r subgroup. (b) Show t h a t a l l C a r t e r subgroups of the solvable group G are conjugate i n G. (c) If C is a C a r t e r subgroup of G a n d G / N is nilpotent for some n o r m a l subgroup N of G, show t h a t N C = G. N o t e . T h i s p r o b l e m and the next one seem quite a bit harder t h a n most of the problems i n this book. W e present t h e m as challenges to the reader, a n d to give a taste of some of the theory of solvable groups t h a t we have omitted. 3 C . 8 . L e t G be solvable. A nilpotent injector of G is a nilpotent subgroup / c o n t a i n i n g the F i t t i n g subgroup F ( G ) , a n d m a x i m a l w i t h this property. P r o v e t h a t a l l nilpotent injectors of G are conjugate i n G. 3D A finite group G is said to be rr -separable, where rr is some set of primes, if there exists a n o r m a l series 1 = N
0

C TVi C c 7V = G
r

such t h a t each factor N . / N ^ x is either a rr-group or a rr'-group. (Observe t h a t a rr -separable group is the same t h i n g as a vr'-separable group.) It is easy to prove, as we shall see, t h a t finite solvable groups are vr-separable for all p r i m e sets rr, and so rr -separability can be viewed as a generalization of solvability. F i r s t , we m e n t i o n t h a t i f G is rr -separable, t h e n so too is every subgroup a n d every h o m o m o r p h i c image. ( T h i s fact is easy to establish, a n d we trust t h a t the reader c a n s u p p l y a proof.) N e x t , we recall t h a t C v ( G ) is the unique largest n o r m a l vr-subgroup i n G , w h i c h may, of course, be the t r i v i a l subgroup. ( T h e uniqueness is an i m m e d i a t e consequence of the fact t h a t i f N

92

3. S p l i t

Extensions

a n d M are n o r m a l rr-subgroups of G, t h e n J V M is also a n o r m a l 7r-subgroup.) Because O ( G ) is u n i q u e l y determined, it is, of course, characteristic i n G.

n

In the definition of rr -separability, we can renumber the n o r m a l subgroups N i i f necessary, i n order to eliminate repeats among t h e m . If G is rr -separable a n d n o n t r i v i a l , therefore, we can assume that N i > 1. Since Ni is either a n o r m a l rr -subgroup or a n o r m a l rr '-subgroup of G, we see I n particular, G has a t h a t either C v ( G ) > 1 or C v ( G ) > 1 (or b o t h ) .

n o n t r i v i a l characteristic subgroup N t h a t is either a rr -group or a rr '-group. A l s o , i f N < G, t h e n G / N is rr-separable, a n d so we can find a n o n t r i v i a l characteristic subgroup M / N of G / N t h a t is either a rr-group or a rr'-group. C o n t i n u i n g like this, we c a n construct a series of characteristic subgroups of G whose factors are rr-groups a n d rr'-groups. The discussion i n the previous paragraph shows that i f we strengthen the c o n d i t i o n on the n o r m a l subgroups N i n the definition of rr-separability, a n d we require t h e m to be characteristic, a n d not merely n o r m a l , t h e n the same class of groups w o u l d result. Somewhat paradoxically, this observation allows us to prove t h a t a weakened version of the definition also yields exactly the same class of groups. It suffices to consider s u b n o r m a l series. 3.18. L e m m a . Suppose that G is a 1 = N i s a series a it'-group. of subgroups Then such
Q

finite

group a n d that
r

< N i< < N

= G

t h a t each

factor A^/A/j-i is either a it-group o r

G is it-separable.

P r o o f . W e c a n assume t h a t r > 0, a n d we proceed b y i n d u c t i o n on r . Since N - x is rr-separable b y the inductive hypothesis, the preceding remarks
T

guarantee t h a t there exists a characteristic series 1 = M

0

C C M
r

= N

- i

h a v i n g rr a n d rr'-factors. Since 7 V _ i < G, the characteristic subgroups M i of 7 V _ i are n o r m a l i n G, a n d hence G has the n o r m a l series
r

1 = M

C C M

= N -x C N
r

= G

with 7 T a n d rr '-factors. T h u s G is rr -separable, as desired. 3.19. every C o r o l l a r y . L e t G be set i t of p r i m e s . finite a n d solvable. Then

G is it-separable f o r

P r o o f . C o n s i d e r a c o m p o s i t i o n series for G. (Recall t h a t this is a c h a i n of subgroups 1 = 7V < N

0 x

< < 7V = G
r

such t h a t each of the factors N / N - i is simple.)

Since we are assuming

t h a t G is solvable, its c o m p o s i t i o n factors are solvable too, a n d thus each of

3D

93

the simple groups N / N - x has prime order. B u t every p r i m e p is either a vr-number or a vr'-number, depending o n whether p G vr or p 0 vr, a n d thus G is vr-separable b y L e m m a 3.18.
t t

A somewhat stronger c o n d i t i o n t h a n "vr-separability" is "vr-solvability". A finite group G is vr -solvable if it has a n o r m a l series where each factor is either a vr'-group or is a solvable vr-group. ( O f course, a solvable group is a u t o m a t i c a l l y vr-solvable for every set vr of primes.) W e s h a l l not have m u c h to say about vr-solvable groups except i n the case where vr = { p } consists of a single p r i m e . Since ^ g r o u p s are solvable, it follows t h a t {p}-separability a n d { ^ - s o l v a b i l i t y are the same t h i n g , and groups i n this class are generally described as being > s o l v a b l e " . N o t e that a finite group is solvable i f a n d o n l y i f it is p-solvable for every p r i m e p . A key result about vr-separable groups is the following, w h i c h is a generalization of H a l l ' s E - t h e o r e m , w i t h essentially the same proof. 3.20. Theorem. L e t G be a I T - s e p a r a b l e g r o u p . Then G has a H a l l vr-

subgroup. P r o o f . W e proceed by i n d u c t i o n o n \ G \ . Suppose first t h a t O ( G )

n

> 1.

B y the i n d u c t i v e hypothesis, the group G / O ( G )

t

has a H a l l vr-subgroup

H/O (G),
n

a n d it is immediate that i f is a H a l l vr-subgroup of G . If C v ( G ) =

1, then since we c a n certainly assume t h a t G is n o n t r i v i a l , we have C v ( G ) > 1. W e can thus a p p l y the inductive hypothesis to G / C v ( G ) to produce a H a l l rr-subgroup H / 0 ^ { G ) . B y the Schur-Zassenhaus theorem, C v ( G ) has a complement i n H , a n d it is easy to see t h a t such a complement is the desired H a l l vr-subgroup of G .

T h e reader might guess t h a t analogs of the C - t h e o r e m a n d D - t h e o r e m also h o l d i n the context of vr-separable groups, a n d indeed they do. T h e i r proofs rely on the conjugacy part of the Schur-Zassenhaus theorem, however, a n d thus they require either some solvability a s s u m p t i o n or else a n appeal to the F e i t - T h o m p s o n theorem. Some of the basic definitions a n d results about vr-separability go back to the work of S. A . C u n i h i n i n 1948, but it was the historic paper of P h i l i p H a l l a n d G r a h a m H i g m a n i n 1956 t h a t established a n u m b e r of deeper properties, at least for p-solvable groups. A basic a n d frequently used result i n the H a l l H i g m a n paper is their L e m m a 1.2.3, w h i c h we state here somewhat more generally: for vr-separable groups. 3.21. assume Theorem
t

( H a l l - H i g m a n 1.2.3). L e t G be a i r - s e p a r a b l e g r o u p , a n d O (G) D C ( O ( G ) ) .
f f g t

t h a t O / ( G ) = 1. Then

94

3. S p l i t

Extensions

P r o o f . W r i t e C = C ( C v ( G ) ) and let B = C n C v ( G ) .
G

O u r goal is to

show t h a t 5 = 0 , a n d so we assume t h a t B < C, and we work to derive a c o n t r a d i c t i o n . N o t e t h a t B a n d C are n o r m a l (and i n fact, characteristic) in G and t h a t B is a vr-group. Since C/B since K / B is a n o n t r i v i a l vr-separable group, it has a n o n t r i v i a l characis either a vr-group or a vr'-group. A l s o , < G / B , we see that K / B < G/B, and is also a vr-group, teristic subgroup K / B , where K / B is characteristic i n C/B

hence K < G . Since ? is a vr-group, it follows t h a t i f K / B

t h e n AT is a n o r m a l vr-subgroup of G , and hence K C C v ( G ) . T h i s is not the case, however, since ? < K C G and 5 = C n C v ( G ) . W e can thus conclude t h a t K / B is a vr'-group. B y the Schur-Zassenhaus theorem, therefore, there is a complement H for B i n K , a n d we see t h a t H > 1. We have H C G = C ( C v ( G ) ) C C ( B ) , where the final containment
G G

holds because 5 C C v ( G ) . T h u s ? centralizes

a n d since B H = K , we

see t h a t H < K . T h u s 1 < H C O ^ ( A T ) < G , a n d this is a c o n t r a d i c t i o n since by hypothesis, O / ( G ) = 1, and hence G has no n o n t r i v i a l n o r m a l
t

rr '-subgroup.

In a group G , a n o r m a l subgroup N t h a t contains its own centralizer can be thought of as a "large" subgroup, and indeed, i n this s i t u a t i o n , the is i s o m o r p h i c a l l y If C ( N )
G

i n d e x | G : N \ , is bounded i n terms of N . To see this, recall t h a t for an a r b i t r a r y n o r m a l subgroup N , the factor group G / C ( N )
G

embedded i n A u t ( / V ) , and so \ G : C ( / V ) | < |Aut(7V)|.

G

C N ,

therefore, we have | G : 7V| < | A u t ( / V ) | . L e m m a 1.2.3 asserts t h a t i n a TTseparable group G for w h i c h C v ( G ) = 1, the n o r m a l subgroup C v ( G ) is large i n this sense. A n o t h e r example of a characteristic subgroup t h a t is guaranteed to be "large" is the F i t t i n g subgroup F ( G ) of a solvable group G. (See P r o b l e m 3B.14 and the note following it.) I n C h a p t e r 9, we discuss the "generalized F i t t i n g subgroup" F * ( G ) , w h i c h always contains its own centralizer, w i t h no e x t r a a s s u m p t i o n o n G . We close this section by showing how the H a l l - H i g m a n L e m m a 1.2.3 can F i r s t , we need a definition. If G is vr-separable, the rr -length be used.

of G is the m i n i m u m possible number of factors t h a t are rr -groups i n any n o r m a l series for G i n w h i c h each factor is either a vr-group or a vr'-group. For example, G has vr-length 1 precisely w h e n G is not a vr'-group and there are vr'-groups is a vr-group. L e t G be a I T - s e p a r a b l e g r o u p , Then a n d suppose that a Hall 1. exist n o r m a l subgroups N C M of G such that N and G / M and M / N 3.22.

Theorem.

I T - s u b g r o u p of G i s a b e l i a n .

the v r - l e n g t h of G i s a t most

P r o o f . C o n s i d e r the vr-separable group G = G / C v ( G ) , a n d observe t h a t Cv(G) = 1. L e t i f be a n abelian H a l l vr-subgroup of G , a n d note t h a t

3E

95

H is a H a l l vr-subgroup of G . _ T h e n H contains every n o r m a l vr-subgroup of G , and i n p a r t i c u l a r , C v ( G ) C H . B u t # is abelian, a n d thus H C C c ( C v ( G ) ) C C v ( G ) , where_the last containment follows by the H a l l H i g m a n L e m m a 1.2.3. T h u s H = 0 * ( G ) , and we have H < G. T h u s 1 C N C T V i i C G is a n o r m a l series for G w i t h just one rr -factor, and the proof is complete.

Problems

3D
p

3 D . 1 . L e t G be a p-solvable group, a n d let P G S y l ( G ) . (a) If C y ( G ) = 1, show t h a t Z ( P ) C O ( G ) .

p

(b) Show t h a t the p-length of G is at most the nilpotence class of P . 3 D . 2 . L e t Z C Z ( G ) and write G = G / Z . Show t h a t C v ( G ) = O ^ G ) for every set rr of primes. 3 D . 3 . L e t G be rr -separable, and assume t h a t C v ( G ) C v ( G ) C Z ( G ) . Show t h a t G is abelian. 3 D . 4 . L e t H act by automorphisms on G , a n d assume t h a t \ H \ and | $ ( G ) | are coprime. Suppose that the induced a c t i o n of # on G / $ ( G ) is t r i v i a l , w h i c h means t h a t each coset of $ ( G ) i n G is setwise invariant under H . Show t h a t the a c t i o n of H on G is t r i v i a l . H i n t . It is no loss to assume t h a t H is a g-group for some prime q. Consider the d e c o m p o s i t i o n of each coset of $ ( G ) into P - o r b i t s . 3D.5. L e t G be p-solvable, and let P e S y l ( G ) . Suppose t h a t P C N ( K ) ,
p G

where K C G is a subgroup of order not divisible by p. Show t h a t i f C Op/(G). H i n t . F i r s t , consider the case where C y ( G ) = I . T o do the general case, consider the group G = G / C y ( G ) .

3E Let A be a group t h a t acts v i a automorphisms o n some group G . B y definit i o n , A permutes the elements of G , b u t because we have a n a c t i o n b y automorphisms, A also permutes m a n y other kinds of group-theoretic objects associated w i t h G . F o r example, A acts o n the set of a l l Sylow p-subgroups of G for any given prime p , a n d A acts on the set of a l l conjugacy classes of G . It is of interest to find, or at least to prove the existence of, ^ - i n v a r i a n t objects of various kinds. Once one has found a n A - i n v a r i a n t object, there are u s u a l l y further actions t h a t c a n be studied. F o r example, i f H is an

96

3. S p l i t

Extensions

A - i n v a r i a n t subgroup of G, then A acts on the set of right cosets or of left cosets of H i n G, a n d on the set of conjugacy classes of H . Of course, the most basic type of A - i n v a r i a n t object i n G is an A invariant element. These A - f i x e d points i n G clearly form a subgroup, w h i c h is denoted C { A ) . T o see w h y this centralizer n o t a t i o n is really quite natu r a l , embed A a n d G i n their semidirect product V = G x A . W e k n o w t h a t the original action of A on G is realized b y conjugation w i t h i n T, a n d thus g = g i f a n d only i f g a n d a commute i n V. T h u s the A-fixed-point subgroup of G is exactly the centralizer of A i n G, a n d this explains the n o t a t i o n C ( A ) , w h i c h is used to denote the fixed-point subgroup even if the semidirect p r o d u c t has not been mentioned. Similarly, w o r k i n g i n the semidirect product, we see t h a t C ( G ) is the set of elements of A t h a t act t r i v i a l l y on G, or i n other words, it is the kernel of the a c t i o n of A on G. T h i s n o t a t i o n also can be used w i t h o u t m e n t i o n i n g the semidirect p r o d u c t .
G a G A

It is especially easy to find A - i n v a r i a n t objects i n G i f A a n d G have cop r i m e orders, a n d this is the s i t u a t i o n we study i n this section. F o r example, we w i l l prove the following. 3.23. T h e o r e m . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d suppose t h a t ( | A | , \ G \ ) = 1. Assume a l s o t h a t a t l e a s t one of A o r G i s s o l v a b l e . Then f o r each p r i m e p , t h e f o l l o w i n g h o l d . (a) T h e r e exists a n A - i n v a r i a n t Sylow p - s u b g r o u p of G. p - s u b g r o u p s of G, then S = T

(b) / / S a n d T a r e A - i n v a r i a n t Sylow f o r some element c C { A ) .

G

N o t e t h a t T h e o r e m 3.23(a) is a generalization of the Sylow E-theorem, to w h i c h it reduces w h e n A is the t r i v i a l group, a n d similarly, 3.23(b) is a generalization of the Sylow C-theorem. A n o t h e r way to t h i n k about this result is to imagine t h a t we have a pair of magic eyeglasses t h r o u g h w h i c h only A - i n v a r i a n t objects can be seen. T h e o r e m 3.23(a) asserts t h a t even w h e n we look at G t h r o u g h these A-glasses, the Sylow E - t h e o r e m can be seen to hold: there exists a v i s i b l e Sylow p-subgroup for each prime p . The Sylow C - t h e o r e m also remains true when G is viewed t h r o u g h the magic A-glasses. W h a t the usual C - t h e o r e m says, of course, is t h a t given two Sylow p-subgroups of G, say S a n d T , we can find an element g G G such t h a t S = T. N o w suppose t h a t S a n d T are Sylow p-subgroups t h a t we can see after d o n n i n g our glasses. (In other words, they are A invariant.) T h e glasses version of the Sylow C - t h e o r e m says t h a t we can see an element g of G such t h a t S = T. T h e assertion, i n other words, is t h a t S a n d T are conjugate b y some A - i n v a r i a n t element: an element of the fixed-point subgroup C ( A ) . (This, of course, is e x a c t l y the assertion
9 9 G

3E

97

of T h e o r e m 3.23(b).) W e w i l l also prove the glasses version of the Sylow D theorem: i f P is a visible (A-invariant) p-subgroup of G , t h e n P is contained in some visible (A-invariant) Sylow p-subgroup. How c a n the glasses version of the Sylow E - t h e o r e m and similar results be proved? In the case where A is a g-group for some p r i m e q, an easy counting argument suffices, as follows. F i r s t , observe t h a t we c a n assume that A is n o n t r i v i a l , and thus q divides | A | , a n d so q does not divide \ G \ . Since | S y l ( G ) | divides \ G \ , it follows t h a t q does not divide | S y l ( G ) | . N o w A acts on S y l ( G ) , and hence this set decomposes into A - o r b i t s , at least one of w h i c h must have size not divisible by q. B u t we are assuming that A is a g-group, and so a l l A - o r b i t s have g-power size. T h e r e must be a n orbit of size 1, therefore, w h i c h means t h a t the set S y l ( G ) contains a n A - i n v a r i a n t member, as required.
p p p p

A similar (but slightly more technical) argument c a n be used to prove 3.23(b) i n the case where A is a g-group, b u t it appears to be u t t e r l y i m possible to make a counting argument like this w o r k if | A | is not a p r i m e power. T h e following result of G . G l a u b e r m a n provides a n effective substitute, however. Unfortunately, the proof of G l a u b e r m a n ' s l e m m a relies on the conjugacy part of the Schur-Zassenhaus theorem, a n d thus we must either appeal to the F e i t - T h o m p s o n theorem, or else assume t h a t something is solvable. (The solvability assumptions t h a t appear i n the statements of T h e o r e m 3.23 and other results about coprime actions are not really necessary, of course, since at least one of | A | or \ G \ is o d d , a n d hence by the F e i t - T h o m p s o n theorem, at least one of A or G must be solvable.) 3.24. L e m m a ( G l a u b e r m a n ) . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d suppose t h a t ( | A | , | G | ) = 1. Assume also that a t l e a s t one of A o r G i s s o l v a b l e . Suppose t h a t A a n d G each a c t o n some n o n e m p t y set Vl, w h e r e t h e a c t i o n of G i s t r a n s i t i v e . F i n a l l y , assume the compatibility condition: (*) (a-g)-a = (a-a)-g
a

f o r a l l a < E V l , a < E A a n d g < E G. T h e f o l l o w i n g t h e n (a) T h e r e exists (b) Ifa,P a-c a n A - i n v a r i a n t element aefi.

hold.

G f l a r e A - i n v a r i a n t , t h e n t h e r e exists = (3.

c G C ( A ) such
G

that

Before proceeding w i t h the proof of G l a u b e r m a n ' s l e m m a , we show how it can be used to prove T h e o r e m 3.23. P r o o f of T h e o r e m 3.23. T h e set S y l ( G ) is n o n e m p t y b y the Sylow E theorem. E a c h of G and A act on S y l ( G ) , where G acts by conjugation and the a c t i o n of A is induced by the given a c t i o n of A on G . (Since each
p p

98

3. S p l i t

Extensions

element of A induces a n a u t o m o r p h i s m of G, it carries Sylow p-subgroups of G to Sylow p-subgroups.) A l s o , by the Sylow C-theorem, the action of G on S y l ( G ) is transitive. Once we check the c o m p a t i b i l i t y c o n d i t i o n of G l a u b e r m a n ' s l e m m a , it w i l l follow b y 3.24(a) that S y l ( G ) contains an A - i n v a r i a n t member, thereby establishing (a). W e need to check that
p p

[s ) =
p

(s y

a a

for S G S y l ( G ) , a G A a n d g G G. (Recall that here, exponentiation by g means conjugation b y g i n G, but exponentiation by a denotes the given action of a . ) Since A acts b y automorphisms, however, we have (S )
9 a

= (g^Sg)*

= {g )- S g

= (S") " ,

as required. A s s e r t i o n (b) is immediate from 3.24(b).

In the previous proof, it was t r i v i a l to check the c o m p a t i b i l i t y c o n d i t i o n of G l a u b e r m a n ' s l e m m a . Indeed, it seems to be equally t r i v i a l to check it i n v i r t u a l l y every a p p l i c a t i o n of 3.24. P r o o f of L e m m a 3.24. T o prove (a), let T = G x > A be the semidirect p r o d uct w i t h respect to the given action of A on G, a n d as usual, identify A and G w i t h appropriate subgroups of T. E v e r y element of T is uniquely of the form a g , where a G A and g G G, and so we can unambiguously define a n a c t i o n of r on ft b y setting a - ( a g ) = { a - a ) - g . T o verify that this really is an action, let a , b G A and g , h G G. T h e n for a G ft, we have (a-{ag))ibh) = (((a-a)- )-6)-/z =
5

{{(a-a)-b)-g )-h a-{(ab)(g h)) a-((ag)(bh)),

b

= =

as wanted, where the second equality holds because of the c o m p a t i b i l i t y c o n d i t i o n , w h i c h we are assuming. Now fix a G ft and let U = T , the stabilizer of a i n T. If x G T, then since by assumption, G is transitive on ft, we can w r i t e a-x = a - g for some element g G G, and thus x g ~ G U. It follows that x G U G , a n d so U G = T. Now G < T , a n d hence U n G < \ U . A l s o ,
a l

\ U : U n G \

= \ U G : G \ = \ T : G \ = \A\

is coprime to \ U D G\, and thus the Schur-Zassenhaus theorem applies i n U w i t h respect to the n o r m a l subgroup U n G. L e t H be a complement for U n G i n U, and note t h a t \ H \ = \ U : U n G\ = \ A \ , and thus H is a complement for G i n T. Since A is also a complement for G i n T, the conjugacy part of the Schur-Zassenhaus theorem allows us to write A = H for some element x G T. B u t H C U, and so H stabilizes a . T h e n A = H
x x

3E

99

stabilizes a-x, and thus a-x is a n A - i n v a r i a n t member of ft. T h i s completes the proof of (a). To prove (b), let a a n d P be A - i n v a r i a n t members of ft. Since G is transitive o n ft, the set X = { g G G \ a - g = 3} is nonempty, a n d we need to find an. A - i n v a r i a n t element i n X . F o r this purpose, we propose to use part (a) of the lemma, w i t h the set X i n place of ft. If x G X , then a-x = /3, a n d so for a A , we have a-x
a

= (a-a)-x

= ( a - x ) - a = (3-a = 3 ,

where the first a n d last equalities h o l d because a a n d 3 are A - i n v a r i a n t , and the second is a consequence of the c o m p a t i b i l i t y c o n d i t i o n . T h u s x G X , a n d so the a c t i o n of A on G maps X into (and hence onto) itself, a n d we have a n action of A on X .
a

Now let H = G p , the stabilizer of 3 i n G. If we a p p l y the above argument w i t h 0 i n place of a and H i n place of X , we conclude t h a t the action of A on G maps H to itself, and this defines a n action v i a automorphisms of A on H . A l s o , since H is a subgroup of G, we see t h a t ( | A | , \ H \ ) = 1 and that at least one of A or H is solvable. Now A acts on H and on X , a n d we show next that H acts t r a n s i t i v e l y on X by right m u l t i p l i c a t i o n . If x G X and h G H , t h e n a-(xh) = (a-x)-h = p - h = P , and thus x h G X a n d H acts on X by right m u l t i p l i c a t i o n , as wanted. I n fact, this a c t i o n is transitive, since if x,y G X , t h e n a r ^ fixes P, a n d so x - ? / G i f , and right m u l t i p l i c a t i o n by x~ y carries x to y .
1 1 l

A t this point we k n o w that A acts v i a a u t o m o r p h i s m s on the group H , t h a t ( | A | , \ H \ ) = 1, and t h a t at least one of A or H is solvable. A l s o , each of A a n d H acts on the nonempty set X , a n d the action of H on X is transitive, and thus we have almost everything we need to a p p l y part (a) of the l e m m a to conclude that X contains an A - i n v a r i a n t element. W h a t remains is to check the c o m p a t i b i l i t y c o n d i t i o n . W h a t we need, i n other words, is ( x h ) = x h for x G X , h G H a n d a G A . T h i s is obvious, however, since A acts on G v i a automorphisms, a n d thus X contains an A - i n v a r i a n t element, as required.
a a a

The magic-eyeglasses version of the S y l o w D - t h e o r e m follows easily from T h e o r e m 3.23(a); its proof does not require a direct appeal to G l a u b e r m a n ' s lemma. 3.25. C o r o l l a r y . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G finite g r o u p s , a n d suppose t h a t ( \ A \ , \ G \ ) = 1. Assume also t h a t a t least of A o r G i s s o l v a b l e . L e t P C G be a n A - i n v a r i a n t p - s u b g r o u p f o r p r i m e p . T h e n P i s c o n t a i n e d i n some A - i n v a r i a n t S y l o w p - s u b g r o u p o are one some fG.

100

3. S p l i t

Extensions

P r o o f . R e p l a c i n g P by a subgroup containing it i f necessary, we can assume t h a t P is not contained i n any s t r i c t l y larger A - i n v a r i a n t p-subgroup of G, and we show t h a t i n this case, P S y l ( G ) . L e t N = N ( P ) , a n d observe t h a t N is A - i n v a r i a n t since it is uniquely determined b y the A - i n v a r i a n t subgroup P . B y T h e o r e m 3.23(a), we can choose an A - i n v a r i a n t member S S y l ( / V ) , and we observe t h a t since P is a n o r m a l p-subgroup of N , it must be contained i n the Sylow p-subgroup S. B y our assumption, however, P cannot be p r o p e r l y contained i n S since S is A - i n v a r i a n t , a n d thus P = S. T h u s P is a S y l o w p-subgroup of its normalizer N , a n d it follows t h a t P is a S y l o w p-subgroup of G. (To see this, let P C Q G S y l ( G ) , and suppose t h a t P < Q. T h e n since normalizers grow i n p-groups, Q n N is a p-subgroup of N s t r i c t l y larger t h a n the Sylow p-subgroup P , and this is impossible.)
p G p p

N e x t , we consider conjugacy classes. In general, if i f is a class of G a n d H C G is a subgroup, t h e n i f n f f m a y be empty, and even i f it is nonempty, all we c a n say is t h a t i f n H is a u n i o n of classes of i f . Usually, i f n H is not a single class of H . (Distinct classes of a subgroup H t h a t are contained in the same class of G are said to be fused i n G.) T h e s i t u a t i o n is under m u c h tighter control if the subgroup is the set of fixed points of a coprime action. I n p a r t i c u l a r , there is no fusion i n t h a t case. 3.26. T h e o r e m . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite groups, and write C = C ( A ) . Suppose that (\A\, \G\) = I , a n d assume t h a t a t l e a s t one of A o r G i s s o l v a b l e . T h e n t h e m a p i f ^ i f n C d e f i n e s a b i j e c t i o n f r o m t h e set of A - i n v a r i a n t c o n j u g a c y classes of G o n t o t h e set of a l l c o n j u g a c y classes of C.
G

P r o o f . L e t i f be a n A - i n v a r i a n t class of G. T h e n A acts on i f , and G acts t r a n s i t i v e l y o n i f by conjugation. T h e c o m p a t i b i l i t y c o n d i t i o n of G l a u b e r man's l e m m a is t h a t ( x ) = ( x ) 3 for elements x G i f , a G A and g G, and as usual, it is t r i v i a l to see t h a t this holds. B y 3.24(a), therefore, K contains a n A - i n v a r i a n t element, and thus i f n C is nonempty. A l s o , by 3.24(b), a l l elements of i f n C are conjugate i n C, and thus i f n C is a class of C.
a a a

We now have the m a p i f ^ i f D C from the set of A - i n v a r i a n t classes of G to the set of classes of C; we must show t h a t it is a bijection. If i f and L are A - i n v a r i a n t classes of G, w i t h i f D C = L C ) C , then since i f D C is n o n e m p t y and is contained i n b o t h i f and L , it follows t h a t i f n L is nonempty, a n d thus i f = L . O u r m a p , therefore, is injective. F o r surjectivity, it is enough to show t h a t every element of C lies i n i f n C for some A - i n v a r i a n t class i f of G. O f course, each element c G C lies i n a unique class i f of G, a n d so it suffices to show t h a t i f is A - i n v a r i a n t . If a G A , t h e n i f is a class of G, and since c = c, we see t h a t c G i f n i f " . T h u s the classes i f and i f are
a a a

3E

101

not disjoint, a n d we conclude that K = K , a n d hence K is A - i n v a r i a n t , as wanted. Now we consider cosets. If A acts on G a n d H C G is A - i n v a r i a n t , t h e n as we observed previously, A permutes the left cosets of H i n G and also the right cosets of H i n G. In the case of a coprime action, we can determine w h i c h left cosets and w h i c h right cosets of H are A - i n v a r i a n t . 3.27. T h e o r e m . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d w r i t e C = C ( A ) . L e t H C G be A - i n v a r i a n t s u b g r o u p , a n d suppose t h a t ( | A | , \ H \ ) = 1 a n d t h a t one of A o r H i s s o l v a b l e . Then t h e A - i n v a r i a n t left cosets of H i n G a n d t h e A - i n v a r i a n t r i g h t cosets of H i n G a r e e x a c t l y those cosets of H t h a t c o n t a i n elements of C.
G

P r o o f . If a G A and c G C, then ( H c ) = H c = H e , and similarly, ( c H ) = c H , and so cosets of H t h a t contain elements of C are A - i n v a r i a n t . Now suppose t h a t X is an A - i n v a r i a n t left coset of H . T h e n H acts t r a n s i tively by right m u l t i p l i c a t i o n on X and, of course, A acts on X a n d A acts via automorphisms on H . T h e c o m p a t i b i l i t y c o n d i t i o n for G l a u b e r m a n ' s l e m m a i n this s i t u a t i o n is [ x h ) = x h for x G X , h G H a n d a G A , a n d this holds since the given a c t i o n of A on G is an a c t i o n v i a automorphisms. B y 3.24(a), therefore, X contains a n element of C, as wanted.
a a a a

To handle the right cosets, it suffices to observe t h a t every right coset of H i n G has the form X " = { x ~ \ x G X } for some left coset X of H i n G. Since X is A - i n v a r i a n t i f a n d only if X " is, the desired result for right cosets is a consequence of t h a t for left cosets. ( A l t e r n a t i v e l y , i f Y is a right coset of H i n G, then y h = h~ y defines a transitive a c t i o n of H on Y, a n d we can a p p l y G l a u b e r m a n ' s l e m m a directly.)
1 l 1 l

A c t i o n s on cosets are especially interesting for n o r m a l subgroups. L e t A act on G v i a automorphisms, and suppose t h a t N < G is A - i n v a r i a n t . O f course, the cosets of N i n G are the elements of the factor group G/N, a n d since these are p e r m u t e d by A , we see t h a t A acts on G/N. I n fact, this is an a c t i o n v i a automorphisms since {NxNy)
a

= {Nxy)

= Nx y

a a

= Nx Ny

{Nx) (Ny) .

The A - i n v a r i a n t cosets of N i n G, therefore, are e x a c t l y the elements of the fixed-point subgroup C ( A ) . W e can use the bar convention to give a clean restatement of T h e o r e m 3.27 i n this s i t u a t i o n , a statement t h a t can be paraphrased as "fixed points come from fixed points (in coprime actions)".
G / N

3.28. finite

Corollary. L e t A a c t v i a a u t o m o r p h i s m s g r o u p s , a n d l e t N < G be A - i n v a r i a n t . Assume

o n G, w h e r e A a n d G a r e t h a t (|A|, \ N \ ) = 1 a n d

102

3. S p l i t

Extensions

t h a t a t l e a s t one of A o r N i s s o l v a b l e .
C C ( A ) =

W r i t i n g G = G/N,
C G T A ) .

w e have

P r o o f . T h e A - f i x e d elements of G are exactly the A - i n v a r i a n t cosets of N , w h i c h by 3.27 are exactly the cosets of N t h a t contain elements of C = C ( A ) . _ B u t the cosets of TV that contain elements of C form the image C of C i n G .
G

3.29. C o r o l l a r y . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e f i n i t e g r o u p s , a n d assume t h a t { \ A \ , \ G \ ) = 1. If t h e i n d u c e d a c t i o n of A o n t h e F r a t t i n i f a c t o r g r o u p G / * ( G ) i s t r i v i a l , t h e n t h e a c t i o n of A o n G i s trivial. P r o o f . W e must show t h a t each element a G A lies i n C { G ) , and so it is no loss to assume t h a t A = (a). S i n c ^ A is cyclic, it is certainly solvable, and thus C o r o l l a r y 3.28 applies w i t h G = G / $ ( G ) . W r i t i n g C = C { A ) ,
A G

we have
G = Cc(A) = C = G$(G),

where the first equality holds since A acts t r i v i a l l y on G . It follows that G = G * ( G ) , and thus G = C by the usual argument. (If C < G , then C is contained i n some m a x i m a l subgroup of G , which necessarily contains $ ( G ) too, a n d this is a contradiction.) 3.30. C o r o l l a r y . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d assume t h a t ( \ A \ , | G | ) = 1 a n d t h a t t h e a c t i o n of A o n G i s f a i t h f u l . T h e n t h e i n d u c e d a c t i o n of A o n G / $ ( G ) i s f a i t h f u l . P r o o f . L e t B = C A ( G / $ ( G ) ) and apply the previous corollary to the action of B on G to deduce that B = l . U G i v e n an a r b i t r a r y group action on a finite set, there is an associated set of positive integers: the orbit sizes. A l t h o u g h there is very little that can be said about the sets of integers t h a t arise i n this way i n general, the s i t u a t i o n is quite different for actions v i a automorphisms on groups, where there is considerably more structure. (See, for example, C o r o l l a r y 8.44.) F o r coprime actions, we can say even more. In t h a t case, it is true that if m a n d n are coprime orbit sizes, then m n must also be an orbit size. T o prove this, we use a remarkable observation of B . H a r t l e y and A . T u r u l l , which allows us to replace the group b e i n g acted on by an abelian group. T h e key step is the next theorem, w h i c h follows fairly easily from results about coprime actions t h a t we have already established. 3.31. T h e o r e m ( H a r t l e y - T u r u l l ) . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d suppose t h a t ( \ A \ , | G | ) = 1. Assume also that

3E

103

a t l e a s t one of A o r G i s s o l v a b l e . T h e n A acts v i a a u t o m o r p h i s m s o n some a b e l i a n g r o u p H i n such a w a y t h a t every s u b g r o u p B C A has e q u a l n u m b e r s of f i x e d p o i n t s o n G a n d o n H . A l s o , t h e r e i s a s i z e - p r e s e r v i n g b i s e c t i o n f r o m t h e set of A - o r b i t s o n G t o t h e set of A - o r b i t s o n H . We extract part of the proof as a separate l e m m a , w h i c h has other

applications. 3.32. L e m m a . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d suppose t h a t ( | A | , \ G \ ) = 1. Assume a l s o t h a t a t l e a s t one of A o r G i s s o l v a b l e . L e t C = C { A ) , a n d suppose t h a t P Syl (G) i s A - i n v a r i a n t . T h e n P n C G Syl (C).
G p p

P r o o f . L e t S G S y l ( C ) , a n d note t h a t S is A - i n v a r i a n t . B y C o r o l l a r y 3.25, some A - i n v a r i a n t Sylow p-subgroup Q of G contains S. B y T h e o r e m 3.23(b), we have Q = P for some element c G C, a n d thus
p c

Pnc

= Q nc

= (Qr] C)

S.

The result now follows because P n C Sylow p-subgroup S .

c

is a p-subgroup of C t h a t contains the

The

following is a very general result about groups a c t i n g on finite sets.

3.33. L e m m a . L e t A be a g r o u p t h a t acts o n finite sets ft a n d A , a n d assume t h a t each s u b g r o u p B C A has e q u a l n u m b e r s of fixed p o i n t s i n ft a n d i n A . T h e n t h e r e i s a b i j e c t i o n f : ft A such that f(a-a) = f(a)-a f o r a l i a e V l a n d a G A . I n p a r t i c u l a r , f maps A - o r b i t s o n ft t o A - o r b i t s o n A , a n d so t h e sets of o r b i t sizes i n t h e t w o a c t i o n s a r e i d e n t i c a l . P r o o f . W e prove the existence of / by i n d u c t i o n on |ft|; the assertions about orbits are then immediate. L e t n e U , where U is an A - o r b i t of smallest possible size i n ft. T h e n | A : A ^ | = \ U \ , a n d b y the m i n i m a l i t y of \ U \ , no subgroup properly containing A ^ can fix a p o i n t i n ft. B y hypothesis, therefore, no such subgroup can fix a point i n A . A l s o b y hypothesis, A fixes some point v G A , and it follows t h a t A is the full stabilizer of v i n A , a n d we have A ^ = A .
M M U

We now define / on U by setting / ( u - x ) = vx for x G A . T h i s is well defined since i f = f i - y , then xy~ G A = A, and thus vx = vy. A l s o , / is injective on U since i f vx = vy, t h e n xy~ G A = A , and thus H - x = n-y. Furthermore, i i a e U , we can w r i t e a = px for some element x G A , a n d hence if a G A , we have
l M l v M

f(a-a) = f{fi'{xa))

= v(xa) = (vx)-a =

f{a)-a.

Now let V = f ( U ) , so that V is the orbit of v a n d | V | = \ U \ . B y t a k i n g B = 1, we see t h a t |ft| = | A | , and thus if U = ft, we have V = A a n d there is

104

3. S p l i t

Extensions

n o t h i n g further to prove. W e can assume, therefore, that U < Cl and V < A , and we observe t h a t A acts on Q - U and on A - V . We w i l l argue t h a t the hypotheses of the l e m m a are satisfied for the actions of A on Q - U a n d A - V . Once this is established, the inductive hypothesis w i l l guarantee the existence of an appropriate bijection from Cl - U to A - V , and this can be used to extend the definition of / to a l l of Cl. If B C A , we wish to show t h a t B has equal numbers of fixed points i n Q. - U a n d i n A - V , and for this purpose, it suffices to show t h a t B has equal numbers of fixed points i n U and V . T h i s is clear, however, since for each element a A and each point a G U, we have f ( a - a ) = f ( a ) - a , and so a fixes a i f and only i f it fixes f { a ) . U

P r o o f of T h e o r e m 3.31. W e proceed i n two steps to prove the first part; the final assertion then follows by L e m m a 3.33. F i r s t , we show t h a t G can be replaced by a nilpotent group, and then we complete the proof by showing t h a t i f G is nilpotent, it can be replaced by an abelian group. For each p r i m e divisor p of \ G \ , choose an A - i n v a r i a n t Sylow p-subgroup of G, and let N be the nilpotent group constructed as the external direct p r o d u c t of the set V of selected Sylow subgroups. Since A acts on P for each member P G V, there is a n a t u r a l action of A on N , and we argue that \ C ( B ) \ = \ C ( B ) \ for each subgroup B C A . Since an element of the direct p r o d u c t N is B - f i x e d if and only i f each of its components is fixed, it is clear t h a t
N G

\C (B)\
N

= Yl
Pev

\ C ( B ) \ ,
P

and thus for each prime p , the p-part of \ C ( B ) \ is equal to \ C ( B ) \ = \ P C \ C \ , where C = C { B ) . B u t P n C G S y l ( C ) by L e m m a 3.32, and thus \ P n C\ is the p-part of \ C \ . T h i s shows that \ C ( B ) \ and \ C \ have equal p-parts for a l l primes p, and hence \ C { B ) \ = | C | , as wanted.
N P G p N N

To complete the proof, we can now assume that G is nilpotent, a n d in p a r t i c u l a r it is solvable w i t h some derived length d. W e proceed by i n d u c t i o n on d. If d = 1, t h e n G is abelian, and there is n o t h i n g to prove. We can thus assume t h a t d > 1, and we define the external direct p r o d u c t K = G' x G / G ' , w h i c h has derived length d - 1. Since G' is characteristic i n G, it is A invariant, and thus A acts v i a automorphisms on G' and on G / G ' , and this induces an action v i a automorphisms of A on K . Since |AT| = \ G ' \ \ G / G ' \ = \ G \ is coprime to A , the inductive hypothesis applies, and thus there exists an abelian group H , acted on by A , a n d such that \ C ( B ) \ = | C K ( - B ) | for every subgroup B C A .
H

3E

105

It suffices now to show t h a t \ C \ = \ C ( B ) \ , where as before, C = C { B ) . U s i n g C o r o l l a r y 3.28 to o b t a i n the second equality below, we o b t a i n
K G

\C (B)\ = \C
K

G / G

,(B)\\C ,{B)\ =
G

\CG'/G'\\CnG'\ \C:CnG'\\cnG'\

= \c\,
a n d the proof is complete. W e can now prove the result about coprime orbit sizes that we mentioned earlier. 3.34. T h e o r e m . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d suppose t h a t ( | A | , \ G \ ) = 1. Assume a l s o t h a t a t l e a s t one of A o r G i s s o l v a b l e . Suppose t h a t t h e r e e x i s t A - o r b i t s of size m a n d n i n G, w h e r e m a n d n a r e c o p r i m e . T h e n t h e r e i s a l s o a n o r b i t of size m n . P r o o f . B y T h e o r e m 3.31, we can replace G by an abelian group of the same order w i t h o u t affecting orbit sizes, and thus we can assume that G is abelian. N o w let X and y be A - o r b i t s i n G of sizes m a n d n , respectively, a n d let x G X a n d y G y. T h e n \ A : A \ = \X\ = m and \ A : A \ = \y\ = n , and since m a n d n are coprime, it follows by L e m m a 3.16 t h a t A = A A and t h a t \ A : A n A \ = m n . N o w write z = xy, a n d observe that A ( ~ ) A C A . We w i l l show t h a t A n A = A , a n d consequently, the size of the A - o r b i t of z is IA : A D A | = m n . W e work i n the next few paragraphs to prove t h a t A Q Ay. S i m i l a r reasoning then shows t h a t A C A , a n d this w i l l complete the proof.
x y x y x y x y z x y z x y z z x

Define the m a p r : G - > G by setting

aeAy

Because G is abelian, the order of the factors i n the product is irrelevant, a n d thus r is well defined. Furthermore, a n d also because G is abelian, r is a group h o m o m o r p h i s m . W e argue t h a t r maps each element of the orbit X into C ( A ) . T o see this, recall that A = A A , a n d so A acts transitively on the set X . If u G X , therefore, then as a runs over the elements of A , each element of X occurs equally often i n the form u . W e conclude t h a t T ( U ) is a power of the product of the elements of the orbit X , and thus r ( u ) is A - i n v a r i a n t , as claimed.
G x y y y a

Let X = ( X ) and Y = ( y ) , so t h a t X and Y are A - i n v a r i a n t subgroups of G. A l s o , since r is a h o m o m o r p h i s m t h a t maps every element of X into C ( A ) , it follows t h a t T ( X ) consists of A - f i x e d elements. L e t Z = X n Y, and note that Z is also an A - i n v a r i a n t subgroup of G. L e t B C A be the (setwise) stabilizer of the coset Zy, and observe t h a t A C B since
G y

106

3. S p l i t

Extensions

A stabilizes b o t h Z and y. Since B stabilizes the coset Zy, it follows by T h e o r e m 3.27 t h a t Zy contains a B - i n v a r i a n t element c, a n d thus y Z c and we can write y = t c for some element i Z .
y

Since c is B - i n v a r i a n t and A C B , we see t h a t r ( c ) = cl l, and so r ( c ) is /^-invariant. A l s o , since t Z C X , we know t h a t r ( i ) is A - i n v a r i a n t , and in particular, it is B - i n v a r i a n t . W e conclude t h a t y \ * \ = r { y ) = r ( i ) r ( c ) is ^ - i n v a r i a n t . Since \ A \ is coprime to \ G \ , however, it follows that y is a power of y l ^ l , and hence y is ^ - i n v a r i a n t and B C A
y A y r

We argue next t h a t A C ?, and thus 4 C i T o see this, note that z = xye Xy, a n d thus Xy = Xz, a n d this coset is stabilized by A . Since y Y, it is easy to check t h a t n Y = ( X n >")y = Z y , and since A stabilizes b o t h Xy and y , we conclude t h a t A stabilizes Zy, a n d so A C ?, as claimed. W e have now shown t h a t A C A , a n d as we remarked earlier, similar reasoning yields A C A . T h i s completes the proof.
z r z 2 z z z y z x

We m e n t i o n that i n the proof of T h e o r e m 3.34, the orbit of z = xy is exactly the set xy. T h i s is easy to see, but this observation does not yield a t r i v i a l proof of the theorem because it is not obvious t h a t \Xy\ = \X\\y\. We close this section w i t h one a d d i t i o n a l remark. G i v e n an arbitrary set TV of integers exceeding 1, one can construct a graph w i t h vertex set M by creating an edge j o i n i n g m and n i n TV whenever m a n d n have a c o m m o n prime divisor. A n immediate consequence of T h e o r e m 3.34 is t h a t i f M is the set of n o n t r i v i a l orbit sizes of a coprime finite group action, then the m a x i m u m distance between any two points i n this graph is 2. F o r a notnecessarily coprime group action, it is s t i l l possible to say something about the corresponding graph: it has at most two connected components. (This theorem of the author and C . Praeger is presented i n C h a p t e r 8,)

Problems 3E 3 E . 1 . L e t A act on G v i a automorphisms, and assume t h a t at least one of A or G is solvable, but do not assume that A and G have coprime orders. If G is n o n t r i v i a l , prove t h a t G contains a n o n t r i v i a l A - i n v a r i a n t p-subgroup for some prime p . H i n t . If P < A is a p-subgroup and C { P ) = 1, show t h a t p does not divide \ G \ . Show i n this s i t u a t i o n that for each prime q, there is an A - i n v a r i a n t Sylow g-subgroup i n G.
G

3 E . 2 . L e t A act v i a automorphisms on G, and assume t h a t ( \ A \ , \ G \ ) = 1 and t h a t at least one of A or G is solvable. Suppose t h a t G = H K , where

3F

107

H a n d K are A - i n v a r i a n t subgroups. Show t h a t C ( C ( ~ ) H ) ( C r \ K ) , C = C {A).

G

where

N o t e . W e a r i n g magic glasses, H and K are visible. T h e fact t h a t G H K means t h a t every element of G is the p r o d u c t of a n element of H w i t h an element of K . T h i s p r o b l e m asserts that this is visible t h r o u g h A-glasses. E v e r y element of G that we can see is the p r o d u c t of a visible element of H w i t h a visible element of K . 3 E . 3 . L e t A act on G v i a automorphisms, a n d let N < G be A - i n v a r i a n t . A s s u m e t h a t [ \ A \ , \ N \ ) = 1 a n d t h a t A acts t r i v i a l l y on b o t h N a n d G/N. Show t h a t A acts t r i v i a l l y on G. H i n t . N o solvability assumption is needed here because it is no loss to assume t h a t A is cyclic. 3 E . 4 . In the usual coprime action s i t u a t i o n w i t h A a c t i n g on G, let C = C { A ) . If H C G is A - i n v a r i a n t , show t h a t \ H : H n G | divides | G : C | a n d t h a t | G : C n B | divides \ G : H \ .
G

H i n t . Use L e m m a 3.32. 3 E . 5 . L e t G be p-solvable, a n d assume t h a t C y ( G ) = 1. L e t P = O ( G ) and w r i t e F = $ ( P ) , so that F < G. N o t e t h a t there is a n a t u r a l action of G on P / F , a n d observe that P is contained i n the kernel of this a c t i o n since P / P is abelian. Show that P is the full kernel of this action.
p

N o t e . I n other words, G / P acts faithfully on P / F , a n d so G / P is isom o r p h i c to a subgroup of A u t ( P / P ) . A l s o , P / F is an elementary abelian p-group, of order p , say, a n d so it is isomorphic to the additive group of an n - d i m e n s i o n a l vector space over a field of order p . It follows that G / P is isomorphic to a group of invertible n x n matrices over this field.
n

3F We close this already long chapter on split extensions w i t h a discussion of certain extensions t h a t are not generally split. G i v e n a group N a n d a positive integer m, we w i s h to construct a l l (up to isomorphism) groups G , where G has a n o r m a l subgroup 7V = N such t h a t G/N is cyclic of order m. A s was the case when we discussed semidirect p r o d u c t s earlier i n this chapter, our "synthesis" w i l l be preceded by some "analysis" that w i l l enable us to understand groups of the type we wish to construct.
0 0

Suppose N < G a n d t h a t G / N is cyclic of order m. L e t g N be a coset t h a t generates the cyclic group G/N, a n d observe t h a t g G N . It should be clear t h a t the element g of N a n d the a u t o m o r p h i s m of N i n d u c e d by g
m m

108

3. S p l i t

Extensions

are relevant to this situation, and i n fact, these two objects, along w i t h the group TV and the integer TO uniquely determine G up to isomorphism. In the following, we w i l l again use the convention t h a t i f N = N , then x N
0

and x 3.35.

GN

are corresponding elements under the given isomorphism. Let G and G

0

Theorem.

be g r o u p s , finite

and letN < G and N

< G ,
0

where G / N and GQ/NQ of G / N a n d G Q / N Q

m

a r e c y c l i c of
0

o r d e r TO. Suppose
0

that n ^

is

an isomorphism from N to N , respectively. = (ft))

a n d l e t g N a n d g N r j be g e n e r a t i n g Assume and that {x )

9 0

cosets

(9 )o

(x )
0

9 0

f o r a l l elements x G N . Then there is a unique i s o m o r p h i s m 9 : G - > G t h a t extends the g i v e n i s o m o r p h i s m N -> 7V a n d t h a t c a r r i e s g t o g .

0 Q

P r o o f . T h e distinct cosets of N i n G are g N w i t h 0 < i < TO, and so each element u G G is uniquely of the form u = g x
l

for some element x G N and

some integer i i n this range. W e are therefore forced to define 9 by setting 9(u) = 9 { g x ) = { g o Y x o , and we must show that 9 is indeed an i s o m o r p h i s m
0 0 i

from G onto G . T h e elements of G

are uniquely of the form ( g o ) x

for

0 < i < TO, a n d it follows easily t h a t 9 is b o t h injective a n d surjective. It remains to prove that 9 is a h o m o m o r p h i s m . W e show first that
m

e(g x)
i

= ( g o Y x o for x e N , and w i t h o u t restriction on the integer i. W r i t e

m

1 = qm + r, where 0 < r < TO. B y hypothesis, g a n d thus { { g ) x )

dtfx) =
m q Q

e N and { g )

{g ) ,
0

= {g )
0

m g

x . W e have
0

9{g {g Yx) = (g ) ((g yx)

r m r m 0
{ i+j j

= (g )
0

m q + r

= (gofxo ,

as wanted. N o w g x g i y = g x y N.
i

for integers i and j a n d elements x , y G it is easy to compute t h a t 9 ( ( g x ) ( g i y ) )

i

Since ( x ^ )

(x )
0

{ 9 o ) 3

9{g x)9{giy). Now

T h i s completes the proof.

suppose t h a t we are given four ingredients: a group TV, a positive

integer TO, an element a G N a n d an a u t o m o r p h i s m a of N . T h e previous theorem guarantees that up to isomorphism, there is at most one group G h a v i n g TV as a n o r m a l subgroup w i t h cyclic factor group G / N of order TO, and h a v i n g generating coset gN, where g But
m

= a and x

= x

for a l l

xN.

given N and TO, not every choice of a G N and a G A u t ( / V ) correm

s p o n d to an actual group. If the group G exists, then since conjugation by g effects the a u t o m o r p h i s m a on N , and a = g a
a

is fixed b y g , we must have

= a . A n o t h e r c o n d i t i o n t h a t clearly must be satisfied is t h a t the m t h A s the following cyclic extension theorem shows, there are no further

power of the a u t o m o r p h i s m a must be the inner a u t o m o r p h i s m induced by a. requirements.

3F

109

3.36. T h e o r e m .

L e t N be a g r o u p a n d TO a p o s i t i v e i n t e g e r , a n d l e t a e N that a
a

a n d a e A u t ( / V ) . Assume

= a

and

= x

f o r a l l x E N . T h e n t h e r e exists a g r o u p G, u n i q u e u p t o i s o m o r p h i s m , h a v i n g N as a n o r m a l s u b g r o u p w i t h t h e f o l l o w i n g p r o p e r t i e s . (a) G / N (b) g (c) x

a m

and

= (gN)

i s c y c l i c of o r d e r TO.

= a. =
x3.

P r o o f . T h e uniqueness follows by T h e o r e m 3.35, and so it suffices to construct G, w h i c h we do by realizing it w i t h i n the s y m m e t r i c group on the set ft of ordered pairs ( i , y ) , where 0 < i < TO a n d y G M . F i r s t , we define an action of N on ft by setting ( i , y ) - x = ( i , y x ) , where 0 < i < TO and x,y G N . It is t r i v i a l to check that this defines a faithful action, a n d so we get a corresponding i s o m o r p h i s m N - > 7V C Sym(ft), where x is the p e r m u t a t i o n effected by x for x G N . F o r convenience, we identify N w i t h N v i a this isomorphism, so that N C S y m ( f t ) .
0 0 0

N e x t , we define a function g : ft -> ft b y setting ( i , = (t + l , y ) for 0 < i < m - 1 and y e N , a n d (TO - 1, y ) = (0, a y ) . Since a = a and cr is a n a u t o m o r p h i s m , we see that ( i , a y ) = (i + 1, a y ) for 0 < i < m - 1, and it follows easily that
a a 5 a ?

(i, y ) g

= ( i , ay )

= { i , a y ) = ( i , y a ) = (i, y ) a

for a l l ( i , y ) Vl. (To check the first equality, observe t h a t each a p p l i c a t i o n of g increases the first component b y 1 m o d u l o m. W e pick up one e x t r a factor of a when the first component changes from TO - 1 to 0, and that factor is unaffected by a d d i t i o n a l applications of the function g . ) T h u s g = a , and i n particular, g is a p e r m u t a t i o n of Q.
m

We show next that x g = gx for a l l x G N . T o verify this, it suffices to check t h a t these functions have the same effect on each element ( i , y ) of ft, a n d so we want ( ( i , y ) x ) g = ((i, y ) g ) x . If 0 < i <TO- 1, we have
a

((i, y ) ) g = ( i , y x ) g = ( i + l , [ y x f )
x

= (i+ l , y x ) = {{i, y)g)x

and also ((TO - l , y ) x ) g = (TO - l , y x ) g = ( 0 , a { y x ) ) = ( 0 , a y x ) = (TO - l , y ) g x

a a a 9 a a

as wanted. It follows t h a t x = x , a n d i n particular, N < G, where G = ( N , g ) . A l s o , G / / V is cyclic, w i t h generating coset g N . A l l t h a t remains to check is that G / N has order TO. Since g = a N , it suffices to show t h a t g i 0 N for 0 < j < TO. B u t ( 0 , 1 ) ^ ' = (j, 1) and
m

110

3. S p l i t

Extensions

(0, l ) s = (0, x ) for x E N , a n d thus g* cannot be an element x N when 0 < j < m. T h i s completes the proof. A s an example of the use of the cyclic extension theorem, we give another construction of the generalized quaternion groups. W e saw these previously, i n P r o b l e m 3A.2, where they were constructed as index-2 subgroups of semidihedral groups. G i v e n a positive integer n , divisible by 8, we start w i t h a cyclic group N of order n / 2 . L e t a be the unique element of N of order 2 a n d let a Aut(AT) be the m a p x x ~ \ so that a fixes a . T a k i n g m = 2, we see that a is the t r i v i a l a u t o m o r p h i s m of N , a n d so it is equal to the inner a u t o m o r p h i s m induced by a . B y the cyclic extension theorem, therefore, the cyclic group N can be embedded as a n o r m a l subgroup w i t h index m = 2 in a group Q , a n d there exists an element g G Q - N such t h a t g = a and x = x ' for a l l x G N . W e argue that every element of Q - N has order 4, a n d thus a is the only element of order 2 i n Q. T o see this, observe t h a t a l l elements of Q - TV have the form gx w i t h x G N , a n d we have ( g x ) = g x g x = g x x = ax~ x = a , which has order 2. T h u s gx has order 4, as claimed. T h e unique group constructed here is denoted Q ; it is the generalized q u a t e r n i o n group of order n . (See the remark concerning nomenclature following P r o b l e m 3A.2.)
m 2 9 1 2 2 9 l n

Problems 3F 3 F . 1 . L e t G be a group of order n divisible by 8, and suppose that N has a cyclic subgroup C of index 2 such t h a t every element of G - C has order 4. Show t h a t G = Q , the generalized quaternion group.
n

3 F . 2 . Show that i n principle, we can construct every solvable group by s t a r t i n g w i t h the t r i v i a l group and repeatedly a p p l y i n g T h e o r e m 3.36. 3 F . 3 . L e t Q = Q . Show that Q has exactly three cyclic subgroups of order 4, a n d t h a t these are t r a n s i t i v e l y permuted by A u t ( Q ) .
8

H i n t . If A and B are cyclic subgroups of Q of order 4, use T h e o r e m 3.35 to construct an i s o m o r p h i s m from Q to Q that carries A to B . 3 F . 4 . L e t S = S L ( 2 , 3), w h i c h is the group of 2 x 2 matrices of determinant 1 over the integers m o d u l o 3. Show t h a t | 5 | = 24 and t h a t it has a n o r m a l Sylow 2-subgroup isomorphic to Q .
8

3 F . 5 . L e t S = 5 L ( 2 , 3 ) as i n the previous problem, a n d note t h a t S has index 2 i n G = G L ( 2 , 3), the full group of invertible 2 x 2 matrices over the integers m o d u l o 3. Show t h a t G-S contains an element g of order 2, but

Problems

3F

111

t h a t it contains no element of order 4. P r o v e t h a t there exists a group H containing S as a subgroup of index 2, where H - S contains an element h of order 4 such t h a t the automorphisms of S i n d u c e d by g i n G and by h i n H are identical. Show t h a t H - S contains no element of order 2 a n d finally, show t h a t b o t h G S and H - S c o n t a i n elements of order 8.

Chapter 4

Commutators

4A We begin w i t h a b i t of review. T h e c o m m u t a t o r of two elements x and y i n a group G is the element x ^ y ^ x y , w h i c h is denoted [ x , y \ . Since [ x , y ] = { y x ) ~ x y , we can t h i n k of the c o m m u t a t o r of x a n d y as the "difference" between yx and xy, and i n p a r t i c u l a r , [ x , y ] = 1 i f and only i f x a n d y commute. A l t e r n a t i v e l y , we can w r i t e [ x , y ] = a r V ' , so [ x , y ] can also be viewed as the "difference" between x and its conjugate x.
l

If H C G and AT C G are subgroups, we w r i t e [ H , K ] to denote the subgroup generated by the set { [ h , k] \ h G H , k G AT} of a l l commutators of elements of J f w i t h elements of K . T h i s subgroup is the c o m m u t a t o r of H a n d AT, and it is t r i v i a l if and only if H and K centralize each other. We stress that, i n general, [ H , K ] is not e q u a l to the set of commutators of elements of H w i t h elements of AT; it is the group g e n e r a t e d by t h e m , or equivalents, it is the unique smallest subgroup of G t h a t contains a l l of these commutators. In p a r t i c u l a r , the subgroup [ G , G ] , w h i c h is generated by a l l commutators i n G, need not consist entirely of commutators. R e c a l l t h a t this subgroup is usually denoted G'; it is the derived subgroup (or c o m m u t a t o r subgroup) of G. T h e derived subgroup G' of G is t r i v i a l if and o n l y i f G is abelian, and as we know, it is the unique n o r m a l subgroup of G m i n i m a l w i t h the property t h a t the corresponding factor group is abelian. T h e r e are a number of useful identities satisfied b y commutators. It is t r i v i a l to check, for example, t h a t [ x , y ] [ y , x ] = 1 for a l l x,y G. I n other words, [ y , x ] = [ x , y ] ~ \ and so i f H , K C G, t h e n the generating commutators of [AT, H] are exactly the inverses of the generating commutators of [ H , K ] , and it follows that [ K , H ] = [ H , K ] . A n o t h e r often-used identity,

113

114

4.

Commutators

w h i c h the reader should check, is t h a t [ x y , z] = [ x , z ] y [ y , z] for a l l x , y , z E G. (In the language of Section 3 B , this says that for each element z E G , the map z] from G to G is a crossed h o m o m o r p h i s m w i t h respect to the conj u g a t i o n action of G on itself.) T h i s identity is the key, for example, to p r o v i n g the following fact. 4.1. L e m m a . L e t H a n d K be s u b g r o u p s normalize [ H , K ] , o r equivalently, [ H , K } < by H a n d K . of a g r o u p G. T h e n H a n d K (H, K ) , t h esubgroup generated

P r o o f . Since [ H , K ] = [AT, H ] , it suffices by s y m m e t r y to prove that H C N { [ H , K } ) . L e t h,x e H and k G K . T h e n [ h x , k ] = [ h , k ] [ x , k ] , and so [h,k] = [hx,k][x,k]e [ H , K ] . In other words, conjugation by x e H maps each of the generators [ h , k ] of [ H , K ] back into [ H , K ] , and thus [ H , K ] C [ H , K ) . It follows t h a t [ H , K ] = [ H , K ) , as wanted. (As usual, we substitute x ' for x to deduce the reverse containment.)
x G x 1 x X 1

Not surprisingly, there is a formula for [ z , x y ] analogous to the identity [ x y , z] = [ x , z ] y [ y , z ] , but it seems pointless to memorize it since it can easily be derived using the formula for [ x y , z ] . T o be specific, we have the following: [z,xy] = [ x y , z ] We
1

= ( [ x , z ] * [ y , z\y

= [y, z \ - \ [ x , z ] ^ ) " = [ z , y ] [ z , x \ y .

have already said t h a t K centralizes H i f and only i f [ H , K ] = 1.

T h i s can be generalized slightly, as follows. 4.2. L e m m a . L e t N be a n o r m a l s u b g r o u p of a g r o u p G, a n d suppose that H , K C G a r e a r b i t r a r y s u b g r o w p s . W r i t e G = G / N a n d f o l l o w t h e standard " b a r c o n v e n t i o n " , so t h a t H a n d K a r e t h e i m a g e s of H _ a n d K i n G u n d e r the canonical h o m o m o r p h i s m G ^ G . Then [H, K] = [H,K], and i n p a r t i c u l a r , H a n d K c e n t r a l i z e each o t h e r i n G if a n d o n l y if [ H , K ] C N . P r o o f . Since overbar is a h o m o m o r p h i s m , we have [ u , v ] = [ u , v ] for arbit r a r y elements u,v G G, and it follows that [ H , K ] = [ H , K \ . is t r i v i a l precisely w h e n [ H , K ] C N . We can also express the relation "AT normalizes H " i n the language of T h i s subgroup

commutators. 4.3. L e m m a . L e t H a n d K be s u b g r o u p s of a g r o u p G. T h e n K C N ( J i ) if a n d o n l y if [ H , K ] C H . I n p a r t i c u l a r , H < G if a n d o n l y if [ H , G] C H .

G

P r o o f . L e t h H and k G AT. Since [ h , k ] = h " ^ , we have h = h [ h , k ] , and thus h G H if and only i f [ h , k] e H . T h e result is now immediate.
k

4A

115

If H and K normalize each other, therefore, t h e n [ H , K ] C H n K . If also H n K = 1 , then [ H , K ] 1, and so H a n d K centralize each other. (We have seen and used this fact previously.) We can use L e m m a s 4.2 and 4.3 to reexamine the n o t i o n of a central series i n a group. R e c a l l that a series of subgroups 1 = H
0

C Hi C C H

= G i C

is a central series i n G i f a l l H i are n o r m a l i n G, and also

Z ( G / H i - i ) for 0 < i < r . B y L e m m a 4.2, the latter c o n d i t i o n can be restated as [ H i , G ] C for 0 < i < r . Conversely, suppose we have a series of subgroups H as above, and assume t h a t [ H G ] C i i ^ i for 0 < i < r . F i r s t , observe that since H i - i C we have [ i ^ , G ] C H and thus L e m m a 4.3 yields H i < G for a l l i . T h e subgroups thus form a n o r m a l series, and indeed they form a central series by L e m m a 4.2.
h U i }

To continue this discussion of central series, it is convenient to define commutators of three or more elements or of three or more subgroups. T h e s t a n d a r d n o t a t i o n a l convention is to use "left association". F o r three elements, for example, the triple c o m m u t a t o r is denned by [ x , y , z ] = [ [ x , y } , z ] , and for three subgroups, the definition is [ X , Y , Z ] = [ [ X , Y ] , Z ] . ( C a u t i o n : in general it is not true t h a t [ X , Y, Z] is generated by the elements of the form [ x , y , z] w i t h x e X , y e Y and z G Z . ) M o r e generally, for n > 2, we define
[ x i , X , . . . , X ]
2 n

[ [ x

, x 2 , . . . , X

] , X

] .

and

s i m i l a r l y for subgroups. suppose t h a t { H i | 0 < i < r } is a central series for G , as before. = [ H , G] C

r

Now Then

[G,G] and thus if r > 2, we have

[ G , G , G ] = [[G,G],G] C [ H - G ]
r U

C P _ ,
r 2

and

i f r > 3, [G, G , G , G] = [[G, G , G], G] C [ F _ , G] C P _ .

r 2 r 3 r f c + 1

C o n t i n u i n g like this, we see t h a t i f k < r + 1, then [G, G , . . . , G] C 7 f _ where there are k copies of G i n the c o m m u t a t o r o n the left.
1 2 3 k k

Clearly, we need some better n o t a t i o n here, a n d so we write G = G , G = [ G , G ] , G = [G, G , G] and i n general, G = [ G ~ \ G ] for k > 1. ( T h i s n o t a t i o n is not completely standard; some authors w r i t e ( G ) to denote our subgroup G , and one can find other notations i n the literature too.) N o t e t h a t G = [G, G] = G', the derived subgroup, but i n general, G w i t h k > 2 is not one of the groups G^" ) of the derived series, and i n p a r t i c u l a r ,
f c 7 k 2 k 1

116

4.

Commutators

G< ) = G" = [ [ G , G ] , [ G , G ] ]

is not usually one of our subgroups G .

4 2 k k

(We

shall see, however, t h a t G" C G , and i n general, G ^ C G \ ) N o t e t h a t the subgroups G are characteristic i n G . In particular, G < G , a n d so by L e m m a 4.3, we have G = [ G , G ] C G . T h e subgroups G , therefore, form a descending series. ( A n o t h e r way to see this is by i n d u c t i o n on k. A s s u m i n g t h a t G C G ~\ we get G = [ G , G] C [ G , G] = G . )
k + l k k f c k k k + l k f c _ 1 fc

R e t u r n i n g now to the subgroups E of onr central series for G , we have G C F _ for 1 < k < r + 1, and i n particular, G C H = 1. N o w consider an a r b i t r a r y nilpotent group, w h i c h , we recall, s i m p l y means t h a t G has a central series. It follows t h a t G = 1 for some positive integer m. M o r e precisely, i f G is nilpotent and has a central series of length r , then Qr+i ( f containments.)
% k r + 1 r f c + 1 0 m = 1 T h e l e n g t h o f a s e r i e g i g t h e n u m b e r 0

Conversely, suppose that G the series 1 = G

k m

= 1 for some positive integer m. Consider

m _ 1

C G

C C G

= G.

(Note t h a t the n u m b e r i n g of these subgroups is "backwards", r u n n i n g from high to low.) Since [ G , G ] = G \ it follows t h a t this is a central series, and thus G is nilpotent. Furthermore, we have seen t h a t G C J f _ , where the H i form an a r b i t r a r y central series i n G . Since the G are contained term-by-term i n the subgroups of an a r b i t r a r y central series, the series of G is called the lower central series of G . T h e length of this series is, of course m - 1, where m is the smallest positive integer such t h a t G = 1.
k + k r f c + 1 k k m

R e c a l l from C h a p t e r 1 t h a t the nilpotence class of a nilpotent group G is the smallest integer r such t h a t G is equal to the t e r m Z of the upper central series of G . W e saw t h a t it is also the smallest integer r such that G has a central series { H i | 0 < i < r } , w i t h H = 1 and H = G. (In other words, the nilpotence class is the length of the shortest possible central series.) W e now k n o w t h a t if G is nilpotent of class r then G = 1. A l s o , it is not possible t h a t G = 1 w i t h m < r + 1 because otherwise, the lower central series of G w o u l d be a central series of length m - 1 < r . T h i s shows t h a t if G is nilpotent of class r , then m = r + 1 is the smallest integer such t h a t G = 1.
r 0 r r + 1 m m

O f course, i f G is an a r b i t r a r y (and not necessarily nilpotent) group, one can still consider subgroups of the form G = [G, G , . . . , G ] , where G occurs m times. B y abuse of n o t a t i o n , the series { G \ i > 1} is often called the "lower central series" of G , even though i n the nonnilpotent case, it is not a true central series because the identity is not one of its terms. If G is finite, then since its lower central series cannot involve infinitely m a n y strict containments, there must be some superscript m such t h a t G = G . It is then immediate that G = G for a l l integers n > m, and this final t e r m of the lower central series of G is denoted G . It is not h a r d to see
m i m m + 1 m n

4A

111

t h a t if_/V < G , then G / N is n i l p o t e n U f a n d o n l y i f G C TV. (To see this, write G = G / N , and observe t h a t G = G by repeated a p p l i c a t i o n of L e m m a 4.2.) It follows t h a t if G is finite, t h e n G is the unique n o r m a l subgroup of G m i n i m a l w i t h the property t h a t the corresponding factor group is nilpotent.
m

A s we mentioned i n C h a p t e r 1, the groups of nilpotence class 1 are exactly the n o n t r i v i a l abelian groups. Somewhat more interesting are groups of nilpotence class 2. F o r these nonabelian groups, we have 1 = G = [G, G , G] = [ G , G], so t h a t the derived subgroup G' C Z ( G ) , or equivalently G / Z ( G ) , is abelian. I n this case, a l l c o m m u t a t o r s are central, a n d hence [ x y , z ] = [ x , z \ y [ y , z \ = [ x , z ] [ y , z \ , and so the m a p [ - , z ] defines a homomorp h i s m from G into G ' . T h e following easy consequence of this is sometimes useful. (Before we state our result, we recall t h a t the exponent of a group G is the smallest positive integer n such t h a t x = 1 for a l l x e G . I n other words, the exponent of a group is the least c o m m o n m u l t i p l e of the orders of its elements.)
3 1 n

4.4. L e m m a . L e t P be a p - g r o u p of n i l p o t e n c e class has $(P) exponent p . C Z(P).

e

2, a n d assume
e

that P ' thus

T h e n t h e e x p o n e n t of P / Z ( P )

divides p . I n particular,

if P ' i s e l e m e n t a r y a b e l i a n , t h e n P / Z ( P ) i s e l e m e n t a r y a b e l i a n , a n d

N o t e t h a t i f P has class 2, then P ' C Z ( P ) , and so b o t h P ' and P / Z ( P ) are abelian. T h u s P ' is elementary abelian if a n d o n l y if e = 1, and to prove t h a t P / Z ( P ) is elementary abelian, it suffices to show t h a t its exponent divides p , and this is the assertion of the l e m m a w h e n e = 1. T h e last conclusion of the l e m m a , therefore, is an i m m e d i a t e consequence of the characterization of the F r a t t i n i subgroup of a p-group as the unique smallest n o r m a l subgroup h a v i n g an elementary abelian factor group. T h i s fact was mentioned previously, but we never gave a formal proof, a n d so we digress to do so now. 4.5. L e m m a . L e t N < P , w h e r e P i s a p - g r o u p . a b e l i a n if a n d o n l y i / $ ( P ) C N . P r o o f . L e t M < P be a m a x i m a l subgroup. Since P is nilpotent, M < P , and we see t h a t P / M has no subgroups other t h a n itself a n d the identity. T h u s P / M has p r i m e order, w h i c h must be p . N o w P / M is abelian, and so P ' C M , and also, given x e P , we have P e M . T h u s P ' and x? lie i n the intersection of a l l m a x i m a l subgroups of G , or i n other words, they lie i n $(P). W e conclude t h a t P / $ ( P ) is elementary abelian, a n d thus also P / N is elementary abelian i f $ ( P ) C TV C P .
X

Then P / N is elementary

Conversely, suppose P / N is elementary abelian, a n d assume t h a t N < P . B y the fundamental theorem of abelian groups, P / N is a direct p r o d u c t of

118

4.

Commutators

n o n t r i v i a l cyclic subgroups, each of w h i c h must have order p . T h e product of a l l but one of these factors has index p , and so is m a x i m a l i n P / N . T h e intersection of those m a x i m a l subgroups of P / N obtained by deleting one of its direct factors of order p is t r i v i a l , and thus N is the intersection of some collection of m a x i m a l subgroups of P . W e conclude t h a t $ ( P ) C TV, as wanted. P r o o f of L e m m a 4.4. T o prove t h a t P / Z ( P ) has exponent d i v i d i n g p , we must show t h a t x ^ Z ( P ) for a l l i e P . W e have [ x P , y ] = [ x , y ] = 1, where the first equality holds since [ - , y \ is a h o m o m o r p h i s m , and of course, the second equality holds because [ x , y ] G P ' , w h i c h has exponent p . T h i s shows t h a t x P commutes w i t h a l l elements y G P , and so x ? G Z ( P ) , as required. B y L e m m a 4.5, the proof is now complete.
e p B e e e e

N e x t , we present a result relating centers and c o m m u t a t o r subgroups. Its proof is based o n a n exception to the general principle t h a t elements of the c o m m u t a t o r subgroup need not be commutators. 4.6. suppose Lemma. L e t A be a n a b e l i a n n o r m a l s u b g r o u p i s c y c l i c . Then G' = [ A , G ] and G' = A / [ A n Z ( G ) ) . I n p a r t i c u l a r , if A i s finite, t h e n \ A \ = \ G ' \ \ A n Z(G)|. and let 9 : A - > A be the of a g r o u p G, a n d

t h a t G/A

P r o o f . Choose a generating coset A g for G / A , then 9(ab) = [ab,g] = [a,g] [b,g]

b

m a p defined by 9 ( a ) = [ a , g ] . (Note that [ a , g] G A since A < G.) If a , b G A , = [a,g][b,g] = 9(a)9(b),

where the t h i r d equality holds since [a,g] and b lie i n the abelian group A . T h u s 9 is a h o m o m o r p h i s m , and C ( g ) . A l s o , 9 ( A ) is a subgroup, and 9 ( A ) C [ A , G] C G'.

ker(0) =

Since A g is a generating coset for G = G / A , we see t h a t J g ) = (g) = G, and thus G = A ( g ) , and it follows t h a t C ( g ) C A n Z ( G ) . T h e reverse inequality is clear, and thus C ( g ) = A n Z ( G ) . T h i s yields 0 ( 4 ) = A / ( A n Z ( G ) ) , a n d so to complete the proof, it suffices to show t h a t 9 ( A ) = G'.

ker(0) =

We already k n o w t h a t 9 ( A ) is a subgroup of G', and so it suffices to show t h a t 9 ( A ) < G and t h a t G / 9 ( A ) is abelian. F i r s t , since A is abelian and 9 ( A ) C A , it is clear that A C Also, 9 ( a ) = [ a , g } 3 = G [ a 9 , g [ = 9 ( a 9 ) G 9 ( A ) for a l l a G A , and thus G It follows t h a t G G = A( ) C and so 0(A) < G , as required. G

N (0(A)).

N (0(A)),

N (0(A)).

Now

(redefining overbars) write G = G / 9 ( A ) , so t h a t G = A j g ) . Since

5

[ a , g ] = [a, ] = 9 ( a ) is t r i v i a l for a l l a G A , it follows t h a t the elements o for

4A

119

a G A together w i t h g form a pairwise c o m m u t i n g set of generators for G, w h i c h is therefore abelian. T h i s completes the proof. A s an example of how L e m m a 4.6 c a n be used, we offer the following. 4 .7. T h e o r e m . L e t A < P , w h e r e P i s a p - g r o u p , a n d suppose that A is a b e l i a n a n d \ A \ = p . Assume that P / A is cyclic, a n d that \A n Z(P)| = p . T h e n t h e n i l p o t e n c e class of P i s TO.
m

P r o o f . L e t Z = A n Z ( P ) , and note t h a t P ' C A a n d \ A : P ' \ = \ Z \ = p by L e m m a 4.6. If TO = 1, then A = Z is central i n P , a n d since P / A is cyclic, it follows t h a t P is abelian, and thus its nilpotence class is 1 = TO, as wanted. We can thus assume t h a t TO > 1, and we proceed b y i n d u c t i o n on TO. Since \ A : P ' \ = p and \ A \ > p , we have P ' > 1, and thus P ' meets Z ( P ) n o n t r i v i a l l y b y T h e o r e m 1.19. T h e n 1 < P ' n Z ( P ) C A n Z ( P ) = Z , and since | Z | = p, we have Z = P ' n Z ( P ) , _ a n d i n p a r t i c u l a r , Z C P ' . N o w w r i t e P = P/Z,_and_observe t h a t ( P ) ' = P ' has index p i n the abelian n o r m a l subgroup A of P . A l s o , ~ P / A = P / A is cyclic, and hence b y L e m m a 4.6, we have \ A n Z ( P ) \ = \ A : ( P ) ' | = p, and thus P satisfies the hypotheses of the theorem. Since \ A \ = p " \ we conclude from the i n d u c t i v e h y p o t h e s i s t h a t P has nilpotence class_m - 1, a n d thus i n the lower central series of P , we have (P) = 1 and ( P ) " " ^ 1. Since P = ( P ) = 1, we have P C Z , and thus P = [ P , P ] C [ Z , P ] = 1.
m 7 1 m m m m + 1 m m

pm

To complete the proof, we must show t h a t P ^ 1. If TO = 2, t h e n _ p i ^ j p t h a t TO > 2, a n d hence P CP =

m _ 1 2 > a n ( g o w e c a n S U p o s e 3 1 1 1 m l m m

P ' C A . W e have P ^ = ( P ) " ^ 1, and thus P " ^ = 4 n Z ( P ) . It follows t h a t P ~ % Z ( P ) , and thus P = [ P - \ P ] ^ 1, as required. Suppose now t h a t P is a nonabelian p-group of order p . A s s u m e t h a t | Z ( P ) | = p, and t h a t P has an abelian subgroup A of index p (which, of course, is necessarily normal). Since P is n o n a b e l i a n , A is n o n t r i v i a l , and thus A n Z ( P ) > 1, and hence Z ( P ) C A . A l s o , P / A is cyclic, and it follows by T h e o r e m 4.7 t h a t that the nilpotence class of P is e - 1. To put this into context, let P be an a r b i t r a r y p-group of nilpotence class r , and order p , where e > 2. Since P has at least one n o r m a l subgroup of index p and the corresponding factor group is abelian, it follows t h a t \ P : P ' \ > p . N o w consider the lower central series
e 2 2 e

P = P
2

> P

> P

> > P
2

= 1,
l m

and observe t h a t \ P : P | = | P : P ' | > p , and of course, \ P : P | > p for 2 < i < r . Then p = | P : P | is the p r o d u c t of r integers, each of w h i c h is at least p, and one of w h i c h is at least p . It follows t h a t e > r + 1, and
e r + 1 2

120

4.

Commutators

thus the nilpotence class r of a p-group P of order p w i t h e > 2 is at most e - l . If it happens t h a t r = e - 1, we say that P has m a x i m a l c l a s s . If \ P \ = P > P and | Z ( P ) | = p and P has a n abelian subgroup of index p , we have seen t h a t P has nilpotence class e - 1, and thus P has m a x i m a l class. I n particular, it is easy to see t h a t d i h e d r a l , generalized quaternion and semidihedral 2-groups satisfy these hypotheses, and so a l l of these groups have m a x i m a l class. I n fact, these are the only m a x i m a l class 2-groups, b u t we w i l l not prove t h a t here. A l s o , we m e n t i o n w i t h o u t proof t h a t for p > 2, there exist m a x i m a l class p-groups that fail to have an abelian subgroup of index p . We close this section w i t h a discussion of the set of elements x (E P such t h a t x = 1, where P is a p-group. I n general, these elements do not form a subgroup. (If P = D , for example, there are exactly six such elements i n c l u d i n g the identity, and of course a set of six elements i n a group of order 8 cannot be a subgroup.) T h e group generated by a l l elements x P such t h a t x P = 1 is denoted fii(P), and more generally, for integers r > 1, one defines ft (P) = ( x G P \ x " = 1). (There does not seem to be any standard name for this characteristic subgroup, which is usually just called "omega-r" of P . )
p 8 p r e 2

If p > 2 and P is a p-group of class at most 2, then the elements x G P such t h a t x = 1 a c t u a l l y do form a subgroup. (Equivalently, S l ^ P ) has exponent d i v i d i n g p i n this case.) A s is shown by D , w h i c h has class 2, the c o n d i t i o n p > 2 is essential. B u t the assumption t h a t the nilpotence class is at most 2 is unnecessarily strong; i n fact, it suffices to assume t h a t the nilpotence class is less t h a n the p r i m e p. T h e proof of this result of P h i l i p H a l l uses "commutator collection", which is the basic idea t h a t underlies our proof for class 2 groups. T h e proof of the full result is considerably more complicated, however, and we do not present it here.
p 8

G i v e n x , y ( E P w i t h P = 1 = y P , we want to show t h a t ( x y ) = 1, and so we investigate how to compute ( x y ) for a r b i t r a r y elements x,y G P and positive integers n . T h e idea here is as follows. W e have
X n

( x y )

= x y x y x y

xy

where on the right, there are n copies of x alternating w i t h n copies of y . If P is abelian, it is t r i v i a l to simplify this. W e can move each copy of x as far as necessary to its left, freely passing t h r o u g h the intervening copies of y , thereby collecting the n copies of x at the left, and leaving n copies of y on the right. T h i s , of course, yields the familiar formula ( x y ) = x y for abelian groups.
n n n

yx

W h a t if P is not abelian? In general, i n any group, i f we have the string and we want to move the x to the left, across the y , we can use the

4A

121

identity yx = x y [ y , x } . I n other words, we c a n s t i l l pass the x t h r o u g h the y , but to do so, we must pay a price. T h e p e n a l t y is t h a t we must insert the c o m m u t a t o r [y, x] to the right of the y . For simplicity, consider the case n = 3. M o v i n g the second x left across the first y (and inserting [ y , x ] a a payment for the privilege of doing this) we have {xyf = xyxyxy = xxy[y, x]yxy .

N e x t , we want to move the t h i r d x leftward i n order to j o i n the first two, and we do this i n three steps, first m o v i n g across y , then across [ y , x ] , a n d finally across another y . T o pay for these transits, we must insert [y, x ] , a n d then [ y , x , x ] , and t h e n another [ y , x ] . W e thus have [xyf = xyxyxy = xxy[y, = = x\yxy

xxy[y,x}xy[y,x}y xxyx[y,x}[y,x,x]y[y,x}y

= x x x y [ y , x] [y, x] [y, x , x } y [ y , x ] y , and at this point, we have collected a l l three copies of x at the left, a n d these are followed by a m i x t u r e consisting of three copies of y , three copies of [y, x ] , a n d one copy of [y, x , x } . We can now continue to "simplify" this, b y collecting the three copies of y t o w a r d the left, to a p o s i t i o n just to the right of the three copies of x . I n order to do this, we w i l l need to move the second y across copies of [y, x] a n d [ y , x , x ] , a n d so we must pay penalties of [ y , x , y ] a n d [ y , x , x , y ] . T o collect the t h i r d y , our penalties w i l l be [ y , x , y ] , [ y , x , x , y ] , a n d also [ y , x , y , y ] a n d [y,x,x,y,y\. W h e n this process is finished, we w i l l have ( x y f = x y w , where w is some c o m p l i c a t e d product of one or more copies of each of [ y , x ] , [ y , x , x ] , [ y , x , y ] , [ y , x , x , y ] and [ y , x , x , y , y ] .
3 3

W e could continue. T h e simplest c o m m u t a t o r appearing as a factor of w is [y, x ] , a n d there are three of these. Suppose we collect these leftward, to a p o s i t i o n just to the right of the three copies of y . T o do this, we w i l l need to pay "transit fees" b y inserting commutators like [ [ y , x , x ] , [y,x]] a n d others. T h i s w i l l result i n a formula of the form ( x y ) = x y [ y , x ] w ' , where w' is a p r o d u c t of commutators of "weight" 3 a n d higher. T h i s seems like quite a mess, but it is possible to do a l l of this i n a systematic way, a n d to keep t r a c k of everything, a n d this is the H a l l " c o m m u t a t o r collection process", w h i c h we w i l l not pursue further.
3 3 3 3

If P has class at most 2, everything becomes m u c h easier. I n t h a t case, the c o m m u t a t o r [y, x] is central, a n d so every copy of [y, x] can be moved to the far right, where it w i l l not interfere w i t h further collection. W e s i m p l y need to count how m a n y copies of [ y , x ] we produce as we collect the n

122

4.

Commutators

copies of x leftward i n ( x y f . T h e leftmost x starts i n its desired p o s i t i o n , and it need not be moved. T h e second x moves across one y , i n c u r r i n g one payment of [ y , x ] ; the t h i r d x moves across two copies of y generating two a d d i t i o n a l copies of [y, x } ; the fourth x generates three more copies of [y, x ] , and so o n . I n t o t a l , therefore, we o b t a i n 1 + 2 + + (n - 1) = (n - l ) n / 2 copies of [ y , x ] , a n d this establishes the formula (xy)
n

x y \ y , x ] ^ - ^ ,

in a nilpotent group w i t h class at most 2. Two useful consequences of this formula are given i n the following result. class a t most 2, and

4.8. T h e o r e m . L e t P be a p - g r o u p w i t h n i l p o t e n c e assume t h a t p > 2 . The f o l l o w i n g then hold. (a) The set { x e P \ x P = 1} i s a s u b g r o u p (b) / / [ y , x ] P = 1 f o r a l l x , y E P , p h i s m f r o m P into itself
p

of P .
x ^ x
p

then

t h e map

isa

homomor-

P r o o f . F i r s t , assume t h a t [y, x } = 1 for a l l x , y E P . Since the nilpotence class of P is at most 2, we c a n a p p l y our c o m m u t a t o r collection formula to conclude t h a t { x y ) = x y P [ y , x ] , where m = p ( p - l ) / 2 . Since p ^ 2, however, m is a m u l t i p l e of p , a n d thus [ y , x } = 1. It follows that ( x y ) P = x P y P , a n d thus the m a p i ^ ^ i s a group h o m o m o r p h i s m , proving (b).
p p m m

Now (without assuming t h a t p t h powers of c o m m u t a t o r s i n P are t r i v i a l ) observe t h a t since a l l c o m m u t a t o r s i n P are central, the m a p [ - , x ] from P to P ' is a group h o m o m o r p h i s m for every element x E P . If y E P a n d
y P = 1, therefore, we have [ y , x }
p m
p

= [y ,x]

= [l,x]

= 1 for a l l x E P.

In

particular, i f b o t h ^ = 1 a n d / = 1, then c o m m u t a t o r collection yields [ x y ) P = x y P [ y , x ] = 1, where as before, m = p ( p - l ) / 2 is a m u l t i p l e of p because p ^ 2. T h e set { x e P \ x P = 1} is thus closed under m u l t i p l i c a t i o n , and so it is a subgroup. Observe t h a t the conclusion of T h e o r e m 4.8(a) is equivalent t o saying t h a t P = 1 for every element x 6
X

fti(P).

Problems 4 A 4 A . 1 . Suppose that G = A B , where A a n d B are abelian subgroups. Show t h a t G' = [ A , B } . 4 A . 2 . L e t H , K a n d L be subgroups of order 2 i n some group G, a n d observe that the set { [ h , k , l ] \ h <E H , k <E K , I <E L } contains at most one n o n i d e n t i t y element, a n d so it generates a cyclic group. N o w let G = A , the
5

Problems 4 A

123

a l t e r n a t i n g group of degree 5. Show t h a t it is possible to choose subgroups H , i f a n d L of order 2 such t h a t [ P , i f , L] = G. H i n t . L e t P C G have order 5. Choose H , i f C N ( P ) w i t h H ^ K , and choose L ^ N ( P ) . Show t h a t [ P , i f ] = P a n d observe t h a t ( P , L ) = G.
G G

N o t e . T h i s p r o b l e m shows that [ P , i f , L] is not necessarily generated by elements of the form [ h , k, 1} w i t h h e H , k e i f a n d / e L . 4 A . 3 . Say that a group Q is q u a s i q u a t e r n i o n i f Q = C U , where C a n d P are cyclic subgroups, w i t h C < Q and C n U = Z ( Q ) . (Note that generalized quaternion a n d semidihedral groups are quasiquaternion.) Suppose t h a t G / Z is quasiquaternion, where Z C G' n Z ( G ) . Show t h a t Z = 1. H i n t . L e t A and B be the subgroups of G = G / Z corresponding to C a n d U of the definition of "quasiquaternion". Show t h a t A n P = Z ( G ) and a p p l y L e m m a 4.6. N o t e . G i v e n any group Q, consider a b e l i a n groups Z t h a t can be isomorp h i c a l l y embedded i n some group G i n such a way t h a t Z C G' D Z ( G ) a n d G / Z = Q. It is a fact t h a t up to i s o m o r p h i s m , there is a unique largest such group Z , depending on Q; it is called the Schur m u l t i p l i e r of Q. T h e h o m o m o r p h i c images of the Schur m u l t i p l i e r are a l l of the groups Z t h a t satisfy our conditions. T h i s p r o b l e m asserts t h a t quasiquaternion groups have t r i v i a l Schur m u l t i p l i e r s . ( A l i t t l e more i n f o r m a t i o n about Schur multipliers appears i n C h a p t e r 5.) 4 A . 4 . A p-group P is said to be e x t r a s p e c i a l i f P ' = Z ( P ) has order p . Show t h a t i f P is extraspecial, then P / Z ( P ) is elementary abelian, or equivalents, Z ( P ) = P ' = $ ( P ) . N o t e . N o n a b e l i a n p-groups of order p
3

are extraspecial.

4 A . 5 . L e t P be a n extraspecial p-group a n d let x , y P w i t h xy + yx. L e t U = ( x , y ) a n d V = C (U).

P P

(a) Show t h a t \ P : C ( x ) \ = p . (b) Show t h a t \ U \ = p . (c) Show t h a t \ P : V \ = p .

2 3

(d) If P > U, show that Z ( V ) = Z ( P ) a n d t h a t V is extraspecial. (e) Show t h a t | P : Z ( P ) | = p H i n t . Use i n d u c t i o n on | P | for e. N o t e . F o r every o d d integer n > 1 and every p r i m e p, there are e x a c t l y two i s o m o r p h i s m types of extraspecial p-groups of order p . If p ^ 2, these groups c a n be distinguished by the fact t h a t one of t h e m has exponent p and the other has exponent p .
n 2 2 e

for some integer e.

124

4.

Commutators

4 A . 6 . L e t P be a p-group h a v i n g m a x i m a l class, a n d let N < P , where \ N : P \ > p . Show t h a t N is one of the terms of the lower central series for P , a n d i n p a r t i c u l a r , N is the only n o r m a l subgroup of G h a v i n g order | W | .
2

4 A . 7 . F i x a p r i m e p a n d let A be the external direct p r o d u c t of p copies of a cyclic group C of order p w i t h n > 0. V i e w the elements of A as p-tuples of the form ( c c , . . . , c ) w i t h a E C. L e t U = ( u ) be cyclic of order p , and let U act on A by c y c l i n g the components. I n other words,
n 1 ; 2 p

(ci,c ,...,Cp-i,c )
2 p

= (c ,ci,c ,.
p 2

. . , C

- l )

F i n a l l y , let P = A x P be the semidirect p r o d u c t , a n d view A a n d U as subgroups of P . Show that the m a x i m u m of the orders of the elements of P is e x a c t l y p .
n + 1

N o t e . O f course, P is just the regular w r e a t h p r o d u c t P = C \ U . n = 1, we get a n example where fti(P) does not have exponent p .

Taking

4 A . 8 . A s s u m e the n o t a t i o n of P r o b l e m 4 A . 7 w i t h n > 0. (a) Show that C ( U ) = Z ( P ) is the set of p-tuples i n A such t h a t a l l

A

components are equal. (b) Show t h a t [ A , U] = P ' is the set of p-tuples i n A such t h a t the p r o d u c t of the components (computed i n C ) is the identity. (c) Show | Z ( P ' P ) | = p . (d) If n = 1, show t h a t P has m a x i m a l class, a n d t h a t i n any case, P ' U has m a x i m a l class. 4 A . 9 . L e t G = N A , where N < G a n d A C G, a n d let M be the final t e r m in the series TV D [ N , A ] D [ N , A , A ] D [ N , A , A , A ] D . (a) Show t h a t A G if and only if M C A .

(b) If M C A , show that M C A. 4 A . 1 0 . L e t P be a p-group w i t h | P : Z ( P ) | < p . Show t h a t |P'| < p ( - )/ .

n n 1 2 n

H i n t . Induct o n n . If P is nonabelian, choose Q so t h a t Z ( P ) C Q C P w i t h \ P : Q \ = p . G e t a b o u n d on \ Q ' \ a n d t h e n a p p l y L e m m a 4.6 to the group P / Q ' . N o t e t h a t { P / Q J = P'/Q'.

N o t e . If the center of a group has s m a l l index, t h e n the group is i n some sense "nearly a b e l i a n " , a n d so it should be no surprise t h a t the derived subgroup is not too b i g .

4B

125

4 A . 1 1 . L e t G = A I H be the regular w r e a t h p r o d u c t of an abelian group A w i t h a n a r b i t r a r y finite group H , a n d view the base subgroup B as the group of functions / : H A w i t h pointwise m u l t i p l i c a t i o n . L e t K C H . Show t h a t [ B , K ] is exactly the set of functions f e B w i t h the property t h a t the product of the values of / over each left coset of K i n H is the identity. Deduce that \ [ B , K } \ = \ A \ \ \ - \ - \ .
H H K

H i n t . T h e action of H o n B is given b y f ( x ) = f { x h ~ )

for h H .

4 A . 1 2 . G i v e n a finite group H , let T ( H ) denote the set of subgroups T of H m a x i m a l w i t h the property that T is abelian a n d c a n be generated by two elements. L e t G = A I H as i n P r o b l e m 4 A . 1 1 , where A is abelian, a n d let B be the base group. (a) Show t h a t if x,y e G a n d [ x , y ] B , then [ x , y ] [ B , T ] for some member T E T ( H ) . (b) Show that i f

t h e n G' D B contains elements t h a t are not commutators i n G. (c) If H is not a n abelian group t h a t c a n be generated b y two elements, show that there exists some number m such t h a t the inequality of (b) holds whenever | A | > m. I n p a r t i c u l a r , show that we can take m = 3 if H = D .
8

4 A . 1 3 . L e t G be nilpotent group w i t h class m > 1, a n d let a G. Show that the nilpotence class of the subgroup H = G ' ( a ) is less t h a n m . 4B

w h i c h holds for every three elements of every group. Its proof, w h i c h proceeds by expansion a n d cancellation, is completely elementary a n d u t t e r l y routine. It is a c o m p u t a t i o n t h a t is fun to c a r r y out, however, a n d we urge the reader to do so. One can t h i n k of the H a l l - W i t t formula as a k i n d of three-variable version of the m u c h more elementary two-variable identity, [ x , y ] [ y , x ] = 1. T h i s observation hints at the p o s s i b i l i t y t h a t a corresponding four-variable formula might exist, but i f there is such a four-variable identity, it has yet to be discovered. We digress briefly to m e n t i o n that the H a l l - W i t t identity is analogous to the somewhat simpler J a c o b i identity for a d d i t i v e commutators i n a ring.

126

4.

Commutators

(We are assuming associativity of m u l t i p l i c a t i o n , of course.) If x,y R, where R is a ring, the additive c o m m u t a t o r of x and y is xy - yx, w h i c h is zero precisely w h e n x a n d y commute. Unfortunately, the s t a n d a r d n o t a t i o n for the a d d i t i v e c o m m u t a t o r of x and y i n a ring is e x a c t l y the same as the n o t a t i o n for group commutators: [ x , y ] . B u t we w i l l not refer to additive c o m m u t a t o r s outside of this paragraph, and so we hope t h a t this a m b i g u i t y w i l l not cause any confusion. T h e J a c o b i identity asserts t h a t [ x , y , z\ + [ y , z , x ] + [ z , x , y ] = 0 for a l l x , y , z R, where R is a ring, a n d where as w i t h groups, the triple commutators are denned b y left association. W e m e n t i o n t h a t a (nonassociative) r i n g L w i t h m u l t i p l i c a t i o n x-y is a L i e r i n g if x - x = 0 for a l l x L a n d the J a c o b i identity ( x - y ) - z + ( y - z ) - x + ( z - x ) - y = 0 holds for a l l x , y , z L . In p a r t i c u l a r , every associative r i n g is a L i e r i n g w i t h respect to the m u l t i p l i c a t i o n x-y = [ x , y ] . In fact, L i e rings provide a powerful t o o l for the study of p-groups, but we w i l l say no more about t h a t here. O u r p r i n c i p a l applications of the H a l l - W i t t i d e n t i t y rely on the following l e m m a a n d its corollary, b o t h of w h i c h are referred to as the "threesubgroups l e m m a . " 4.9. L e m m a . L e t X , Y a n d Z be s u b g r o u p s of a n a r b i t r a r y g r o u p G, a n d suppose t h a t [ X , Y, Z] = 1 a n d [ Y , Z , X ] = 1 . T h e n [ Z , X , Y ] = 1 .

P r o o f . W e want to show t h a t [ [ Z , X ] , Y ] = 1, or equivalently, that every element of the group [ Z , X ] commutes w i t h every element of Y. F o r this purpose, we show t h a t a l l commutators [ z , x] for z Z a n d x X centralize each element y Y. T h i s is sufficient because C ( y ) is a subgroup of G, and so i f it contains a l l of the commutators [ z , x ] , then it contains the subgroup [ Z , X ] generated by these commutators. It is, therefore, enough to show t h a t [ z , x , y ] = 1, for a l l x X , y Y and z Z . ( B u t recall t h a t i n general, these commutators do not generate [ Z , X , Y ] . ) Equivalently, it suffices to show t h a t [ z , x ' \ y ] = 1 for a l l x , y and z i n X , Y a n d Z , respectively.
G

N o w [ x ^ y - ] [ X , Y ] , a n d so [ x , y - \ z ] [ X , Y , Z ] = 1, and we have [ x , y ' \ z } y = 1. Similarly, [ y , - \ x ] = 1, and thus by the H a l l - W i t t identity, [ z , x - \ y } = 1. T h u s [ z , x - \ y ] = 1, as wanted.

z Z x

4.10. C o r o l l a r y . L e t N be a n o r m a l s u b g r o u p of a g r o u p G, a n d l e t X , Y , Z C G be a r b i t r a r y s u b g r o u p s . I f [ X , Y , Z ] C N a n d [ Y , Z , X ] C N , t h e n [ Z , X , Y] C N. P r o o f . L e t G = G / N a n d follow the s t a n d a r d L e m m a 4.2, we have [ H j C \ Then [X,Y,Z] = [[X,Y],Z] = [\X7Y],Z\ = [X,Y,Z] = 1 = [H,K] "bar convention". By

for a l l subgroups H and K of G.

4B

127

and and

similarly, [ Y , Z , X ] = 1. B y L e m m a 4.9 then, 1 = [ Z , X , Y] = [ Z , X , Y] thus [ Z , X , Y ] C N . U

A n appeal to the three-subgroups l e m m a i n a proof often makes it possible to avoid tedious and messy elementwise c o m m u t a t o r calculations. U n fortunately, it is not always easy to see how to a p p l y this very useful tool. We give a number of applications here a n d i n what follows, a n d these s h o u l d provide plenty of practice. F i r s t , we o b t a i n some a d d i t i o n a l i n f o r m a t i o n about the lower central series. 4.11. T h e o r e m . As u s u a l , w r i t e G t o d e n o t e t h e n t h t e r m of t h e l o w e r c e n t r a l series of a g r o u p G. T h e n [ G \ GP] C G i f o r i n t e g e r s i , j > 1.
i + n

P r o o f . W e proceed b y i n d u c t i o n on j , w h i c h is the superscript o n the right in the c o m m u t a t o r [ & , & ] . Since G = G, we see that if j = 1, the formula we need is [ G \ G ] C G , a n d this is v a l i d since b y definition, G = [ G \ G ] . W e can assume, therefore, t h a t j > 1, a n d so we can w r i t e 0 = \ G i - \ G \ . Then
l i + 1 i + l

[ G \ G ] = [ G , G<] = [ G ~ \ G, G*], and to show t h a t this t r i p l e c o m m u t a t o r is contained i n the n o r m a l subgroup G

i +

i , it suffices by the three-subgroups l e m m a ( C o r o l l a r y 4.10) to show that [G, G \ G ' } C G

j 1 i + j

and

[ G \ G ~\

G] C G

i + j

We

have [G, G\ G ~ ] = [ G \ G
j l

[ G

\G ~ } C

= G

i + j

where the containment is v a l i d by the i n d u c t i v e hypothesis. A l s o , [G\ G - \ G ] C [ G

j i + j

- \ G] = G

i + j

where here too, the containment follows from the i n d u c t i v e hypothesis. T h i s completes the proof.

Suppose we consider commutators of n copies of G t h a t are not necessarily left associated. F o r example, i f n = 8, we m i g h t have something like [[[G, G], [G, G , G]], [[G, G], G } } . (We refer to such a n object as a c o m m u t a t o r of weight n.) T h e following result shows t h a t the left-associated weight n c o m m u t a t o r of copies of G contains every other such weight n c o m m u t a t o r . (Of course, the left associated weight n c o m m u t a t o r is e x a c t l y G , the n t h t e r m of the lower central series.)
n

4.12. C o r o l l a r y . L e t G be a g r o u p . copies o f G is contained i n G.

n

T h e n every

w e i g h t n c o m m u t a t o r of

128

4.

Commutators

P r o o f . W e proceed b y i n d u c t i o n on n . T h e only weight 1 c o m m u t a t o r is G itself, a n d so there is n o t h i n g to prove i n this case. A n a r b i t r a r y weight n c o m m u t a t o r w i t h n > 1 has the form [ X , Y ] , where X is a weight i c o m m u t a t o r , Y is a weight j c o m m u t a t o r a n d i + j = n . Since i < n a n d j < n , the inductive hypothesis yields X C G a n d Y C G > , and thus [ X , Y] C [ G \ G ' } C G ^ ' = G by T h e o r e m 4.11.
i i + n

Consider the terms G^ of the derived series of G. We see [ G , G] is a weight 2 commutator, G" = [ G ' , G'] = [ [ G , G ] , [ G , G]] 4 c o m m u t a t o r , a n d G'" = [ G " , G " } is a weight 8 c o m m u t a t o r . G is a weight 2 c o m m u t a t o r for a l l r > 0, a n d this yields the
r 2

that G' = is a weight In general, following.

4.13. C o r o l l a r y . We h a v e G C G " f o r a l l g r o u p s G a n d i n t e g e r s r > 0. A l s o , if G i s m l p o t e n t of class m, t h e n t h e d e r i v e d l e n g t h of G i s a t most l + log (m).

2

P r o o f . Since G^ is a weight 2 c o m m u t a t o r , the first assertion is immediate from C o r o l l a r y 4.12. N o w assume t h a t G is nilpotent of class m , a n d let d be its derived length, so t h a t G^ = 1 a n d d is the smallest integer w i t h this property. W e have
r

G "' a n d so 2 ~
d l

D G^ ~
d l

l)

> 1=

,
2

< m + 1. T h e n 2 ~

< m, a n d hence d - 1 < l o g ( m ) .

N e x t , we explore some connections between the conjugacy class sizes and the nilpotence class of a nilpotent group G. Suppose we k n o w the number m of different sizes of conjugacy classes i n G. If m = 1, for example, then since the size of the class of the identity of G is 1, it follows that a l l classes of G have size 1, a n d so G is abelian. ( O f course, we d i d not use the assumption that G is nilpotent here.) T h i s suggests the (still unresolved) question of how the number m of conjugacy class sizes of G affects the structure of G. In p a r t i c u l a r , it w o u l d be interesting to know i f the derived length of G, or perhaps even the nilpotence class of G, is b o u n d e d by some function of m. After m a n y years w i t h o u t progress o n these questions, it was proved by K . Ishikawa t h a t if m = 2, then G has nilpotence class at most 3. (In this case too, it is not really necessary to assume t h a t G is nilpotent; a result of N . Ito shows t h a t a finite group w i t h just two class sizes is necessarily nilpotent.) M o r e recently, A . M a n n proved a stronger result w i t h a n easier proof, and we present t h a t next. G i v e n a finite group G w i t h class sizes 1 = m < n < < n , we write M { G ) to denote the subgroup of G generated by a l l elements l y i n g in conjugacy classes of size at most n . (We w i l l use this n o t a t i o n for the
2 m 2

4B

129

remainder of this section.) N o t e t h a t M ( G ) = G i f m = 2, a n d thus the following result contains Ishikawa's theorem. 4.14. T h e o r e m ( M a n n ) . If G i s a class o / M ( G ) i s a t most 3. finite n i l p o t e n t g r o u p , then the mlpotence

A c t u a l l y , we can replace the hypothesis t h a t G is nilpotent i n T h e o rem 4.14 b y a m u c h weaker one: t h a t G has a self-centralizing n o r m a l subgroup. ( B y this, we mean a subgroup A < i G such t h a t A = C { A ) . Observe t h a t such a subgroup is a u t o m a t i c a l l y abelian.)
G

4.15. T h e o r e m . n o r m a l subgroup. most 3.

Suppose Then

t h a t a finite g r o u p G c o n t a i n s a s e l f - c e n t r a l i z i n g M ( G ) i s n i l p o t e n t , a n d i t s n i l p o t e n c e class is a t

R e c a l l t h a t a nilpotent group G always has a self-centralizing n o r m a l subgroup, a n d so T h e o r e m 4.14 is i m m e d i a t e from T h e o r e m 4.15. ( T h e existence of self-centralizing n o r m a l subgroups i n nilpotent groups is part of P r o b l e m I D . 1 0 , but nevertheless, we give the easy proof here.) 4.16. L e m m a . L e t G be a finite n i l p o t e n t g r o u p , a n d l e t A be m a x i m a l a m o n g n o r m a l a b e l i a n s u b g r o u p s o f G . Then A = C { A ) , a n d hence every nilpotent group contains a self-centralizing n o r m a l subgroup.
G

P r o o f . W r i t e G = C ( A ) . T h e n A C C, a n d we assume t h a t this containment is proper. Since A o G , we have C < G, a n d thus we c a n choose a subgroup B C C such t h a t B / A is m i n i m a l n o r m a l i n G / A . Since G/A is nilpotent, it follows that B / A C Z ( G / A ) , a n d thus B / A has p r i m e order. In p a r t i c u l a r , B / A is cyclic, a n d since B C G = C ( A ) , it follows t h a t A C Z ( B ) . W e conclude that B is abelian, a n d since B < G , this contradicts the choice of A . U
G G

The

key ingredient i n the proof of T h e o r e m 4.15 is the following.

4.17. L e m m a . L e t K be a n a b e l i a n n o r m a l s u b g r o u p of a finite g r o u p G, a n d l e t x E G be n o n c e n t r a l . Then \ C ( x ) \ < \ C ( y ) \ , where y = [k,x] and k E K is arbitrary.

G G

P r o o f . Observe that y < E [AT, G] C AT. If y = 1, t h e n C { y ) = G >

G

C (x),
G

a n d there is n o t h i n g further to prove, so we assume that y ^ 1. B o t h x a n d y lie i n the subgroup H = K C ( x ) , a n d of course, C { x )

G H

= C (x)
G H

and we

C ( v ) Q
H

C { y ) . It suffices, therefore, to prove t h a t | C ( a ; ) | < \ C ( y ) \ .

G H

Now have

let 6 : K

AT be the m a p defined by 6 { t ) = [ t , x ] . F o r s,tK,

6(st) = [st,x] = [ s , x ] % x ]

= [s,x][t,x]

e(s)6(t),

130

4.

Commutators

where the t h i r d equality holds because b o t h t a n d [ s , x] lie i n the abelian group AT. T h u s 6 is a h o m o m o r p h i s m , a n d 6 ( K ) is a subgroup, a n d we have \ 9 { K ) \ = \ K : ker(0)| = \ K : K n C { x ) \ = \ H : C { x ) \ .
H H

We argue next that each of the subgroups AT and C ( x ) normalizes B ( K ) , a n d thus B ( K ) < K C ( x ) = H . C e r t a i n l y , AT C N ( 0 ( A T ) ) since 9 { K ) C K a n d AT is abelian. A l s o , since 0(AT) is uniquely determined by the n o r m a l subgroup AT a n d the element x , we see that C ( x ) C N ( 9 ( K ) ) . Now 0(AT) < H , a n d y is a n o n i d e n t i t y element of 6 { K ) , and thus the entire i f - c o n j u g a c y class Y of y consists of nonidentity elements of 6 { K ) . T h e n |JJ : C ( y ) \ = | ^ | < |#(AT)| = \ H : C { x ) \ , as required.
G G G G G H H

4.18. C o r o l l a r y . L e t K be a n a b e l i a n n o r m a l s u b g r o u p of a n a r b i t r a r y g r o u p G, a n d w r i t e M = M ( G ) . Then [AT, M ] C Z ( G ) .

finite

P r o o f . W r i t e G = G / Z ( G ) . It suffices to show t h a t [AT~, M ] = 1, a n d for this purpose, it is enough to check that x centralizes K for a l l elements x i n some generating set for M . It suffices, therefore, to observe that [ k , x ] e Z ( G ) for a l l k <E K , where x lies i n a conjugacy class of size 1 or n i n G. O f course, i f x e Z ( G ) , there is n o t h i n g to prove, and if x lies i n a class of size n , then b y L e m m a 4.17, the element [ k , x ] lies i n a smaller class, and hence [ k , x ] e Z ( G ) , as required.
2 2

P r o o f of T h e o r e m 4.15. L e t K < G be self-centralizing, and write M = M(G). B y C o r o l l a r y 4.18, we have [AT, M ] C Z ( G ) , a n d thus [AT, M , M ] = 1 a n d [ M , A T , M ] = 1. B y the three-subgroups l e m m a , [ M , M , K ] = 1, a n d thus [ M , M ] C C ( A T ) = AT. T h e n M = [Af, M , A f , Af] C [ K , M , M ] = 1, a n d it follows t h a t A f is nilpotent w i t h class at most 3.
4 G

We clearly cannot drop the assumption i n T h e o r e m 4.15 t h a t G contains a self-centralizing n o r m a l subgroup since, for example, i f G is nonabelian and simple, t h e n M ( G ) = G is certainly not nilpotent. W e do, however, o b t a i n the following very general result. 4.19. M Theorem. L e t G be a the nilpotence finite class group, and w r i t e F = F(Af), 4.
n

where

= M ( G ) . Then

of F i s a t most

P r o o f . L e t n be the nilpotence class of F , so t h a t F > 1 a n d F A s s u m i n g , as we may, t h a t n > 3, consider the characteristic F ~ of G . If this subgroup is abelian, then by C o r o l l a r y 4.18, [ F ~ , M ] C Z ( G ) , a n d thus F = [ F " , F , F ] C [ F " , M , F ] = is a c o n t r a d i c t i o n . W e conclude t h a t F ~ is nonabelian, a n d 1 < [F ' ,F ' } C F by T h e o r e m 4.11. B u t F = 1, 2n - 4 < n + 1. T h i s yields n < 4, as wanted.
n 2 n 2 n n 2 n 2 n 2 n 2 n 2 2 n 4 n + l

= 1. subgroup we have 1, w h i c h therefore and thus

4C

131

Problems 4B 4 B . 1 . L e t G be nilpotent of class exceeding 2. Show t h a t G has a characteristic abelian subgroup that is not central. 4 B . 2 . L e t G be nilpotent. Show t h a t G has a characteristic subgroup K such t h a t K D C ( K ) a n d the nilpotence class of K is at most 2.
G

H i n t . Use P r o b l e m 4 B . 1 . 4 B . 3 . L e t { Z j \ j > 1} be the upper central series for an a r b i t r a r y group G. Show t h a t [ G \ Zj] C Z j - i for a l H > 0 a n d a l l j . I n p a r t i c u l a r , [ G \ Z \ = 1 for i > 0.
{

4 B . 4 . L e t X , Y C G be a r b i t r a r y subgroups, a n d suppose t h a t Y centralizes [X,Y]. (a) P r o v e t h a t Y' centralizes X . (b) N o w assume that X < G. Show t h a t [ X , Y) is abelian. 4 B . 5 . L e t G be supersolvable. Show t h a t M ( G ) is nilpotent w i t h class at most 3. 4C Let A act on G v i a automorphisms. Since A and G c a n be viewed as subgroups of the semidirect product r = G x A , the c o m m u t a t o r [ G , A ] can be c o m p u t e d as a subgroup of T. B u t G < T, a n d so [G, A] C G , a n d i n fact it is possible (and often convenient) to t h i n k of [G, A ) as being c o m p u t e d entirely w i t h i n G , thereby rendering the semidirect product irrelevant. T o do this, observe t h a t i f g e G and a G A , t h e n [ ,a] = ^ a ^ a = $"V, where we c a n view # either as the result of conjugating g by a i n T, or alternatively, as the result of a p p l y i n g a to g i n the o r i g i n a l action. T h e c o m m u t a t o r [G, A ] , therefore, is the subgroup { g ~ g \ g < E G , o G /4), and this w o u l d make sense even i f we d i d not k n o w about semidirect products. Nevertheless, to prove facts about [G, A ] and related objects, we w i l l appeal to semidirect products when it is convenient to do so, and i n particular, we w i l l a p p l y the three-subgroups l e m m a to subgroups of the semidirect product.
5 a l a

W e can a p p l y L e m m a 4.1 (in the semidirect p r o d u c t ) to deduce t h a t [G, A ] < G. A l s o A normalizes [G, A ) i n G x A , w h i c h means t h a t [G, A ] is invariant under conjugation by elements of A , or equivalently, it is invariant under the original action of A on G . I n general, i f H C G is an a r b i t r a r y A - i n v a r i a n t subgroup, we have a n a t u r a l action of A on H , and sometimes i n this context, we say that H a d m i t s the a c t i o n of A . I n particular, since

132

4.

Commutators

[ G , A ] admits the a c t i o n of A , we can compute [ G , A , A ] , w h i c h is n o r m a l in [ G , A ] , but not necessarily i n G. (It is, of course, s u b n o r m a l i n G.) A l s o , [ G , A , A ] admits A , a n d thus we can compute [ G , A , A , A ] , and we can continue like this. A l l of these repeated commutators, of course, a d m i t A and are s u b n o r m a l i n G. R e c a l l t h a t i f A acts v i a automorphisms on G and H C G is A - i n v a r i a n t , then A permutes the set of right cosets of H i n G a n d also the set of left cosets of H i n G. I n p a r t i c u l a r , i f H < G, we have an action of A on G / H , and it is easy to see that this is an action v i a automorphisms. ( T h i s is sometimes called the i n d u c e d a c t i o n of A_on G / H . ) F o r g G G a n d a G A , we have ( g H ) = g H , a n d so if we write G = G / H we have ( g ) = g , and hence
a a a 5

[9, a] = g ' g The

= ( g - ) ^ ) = (g)~H9)

= foa]

It follows t h a t [ G , A ] = [ G , A \ . following characterization of [ G , A ] is sometimes useful.

4 .20. L e m m a . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e g r o u p s . T h e n [ G , A ] i s t h e u n i q u e s m a l l e s t A - i n v a r i a n t n o r m a l s u b g r o u p of G such t h a t t h e i n d u c e d a c t i o n of A o n t h e f a c t o r g r o u p i s t r i v i a l . P r o o f . Suppose t h a t N < G, where N is A - i n v a r i a n t , and w r i t e G = c o m p u t a t i o n is equivalent to 1 = [ G , A ] . G/N.

T h e n A acts t r i v i a l l y on G / N if and only i f 1 = [ G , A ] , w h i c h by the previous T h i s , of course, is the same as [ G , A ] C N , and the result is now immediate.

4 .21. C o r o l l a r y . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e groups, and l e t H Q G . (a) E v e r y r i g h t coset (b) E v e r y left coset Then the f o l l o w i n g are equivalent. of H i n G i s a n A - i n v a r i a n t of H i n G i s a n A - i n v a r i a n t subset. subset.

(c) [ G , A ] C H . I n p a r t i c u l a r , [ G , A] i s t h e u n i q u e s m a l l e s t s u b g r o u p of G w i t h t h e p r o p e r t y t h a t a l l of i t s r i g h t cosets a r e A - i n v a r i a n t , a n d s i m i l a r l y if " l e f t " r e p l a c e s "right". P r o o f . F i r s t , observe that i f X is a subset of G and we write X ~ = { x ' \ x G X } , then the m a p X ^ X ~ takes the right cosets of H to the left cosets of H and v i c e v e r s a . It follows that an a u t o m o r p h i s m of G fixes a l l right cosets of H i f a n d o n l y i f it fixes a l l left cosets of H . W e see, therefore, t h a t (a) and (b) are equivalent.
l l 1

Now assume (b), so that every left coset of H is A - i n v a r i a n t . T h e n for g G G and a G A , we have g G ( g H ) = g H , and so [g, a] = g ' g G H and (c) follows.
a a 1 a

4C

133

F i n a l l y , assume (c), so t h a t [ G , A ] C H . E a c h right coset of H i n G , therefore, is a u n i o n of right cosets of [G, A ] i n G . B u t a l l cosets of [G, A ] in G are A - i n v a r i a n t b y L e m m a 4.20, a n d so unions of such cosets are also A - i n v a r i a n t , and (a) follows. R e c a l l that an action of a group A on a set ft is faithful i f the i d e n t i t y of A is the only element that fixes every p o i n t . I n general, the kernel of an a r b i t r a r y action of a group A is the (necessarily normal) subgroup B of A consisting of the elements of A that act t r i v i a l l y . I n this situation, a l l elements of each coset B a of B i n A mduce the same p e r m u t a t i o n on ft, a n d so we have a well denned action o f A = A / B on ft, where a - a = a - a . T h i s action of A / B is clearly faithful. Now suppose that A acts on G v i a automorphisms. W o r k i n g i n the semidirect p r o d u c t G x > A , we see that the kernel of the action of G on A is C { G ) . ( A s we remarked previously, it is c o m m o n to w r i t e C ( G ) to denote the kernel of the action of A on G even if we do not e x p l i c i t l y construct the semidirect product.) T h i s kernel is the largest subgroup B C A such t h a t [ G , B ] = 1. If B = C { G ) , then B < A , a n d w r i t i n g A = A / B as before, we see t h a t the n a t u r a l action of A on G is given by g = g for g < E G and a E A , a n d it follows t h a t a subgroup H C G is A - i n v a r i a n t if a n d only if it is A - i n v a r i a n t . (In other words H admits A if a n d only i f it admits A . ) A l s o , [ g , a \ = [g, a] for g G and a G A , and so if H admits A and A , we have [ H , A ] = [ H , A \ .
A A A u a

W h a t can we say about the structure of a group A i f we k n o w t h a t it acts via automorphisms on some group G i n such a way t h a t [ G , A , . . . , A ] = 1, where there are (say) m copies of A i n this c o m m u t a t o r ? T h e answer is t h a t we can say precisely nothing. T h i s is because an a r b i t r a r y group A can act t r i v i a l l y on G , i n w h i c h case [G, A ] = 1. M o r e generally, i f B is the kernel of the action of A on G , then B and G cannot "see" each other, a n d so the fact t h a t [ G , A , . . . , A ] = 1 tells us n o t h i n g about B . T o draw a conclusion about the structure of A from some assumption about how G sees the action of A , we must therefore assume that the action is faithful. To facilitate the discussion, we introduce some (nonstandard) n o t a t i o n . We w r i t e [ G , A , . . . , A ] to denote the c o m m u t a t o r w i t h exactly m copies of A . T h e n [ G , A , . . . , A ] i is s i m p l y [G, A ] and i n general [[G, A ] , A , . . . , A ] = [G,A,...,A] .
m m m + 1

4 .22. T h e o r e m . L e t A a n d G be g r o u p s . Suppose t h a t A acts faithfully o n G v i a a u t o m o r p h i s m s , a n d assume t h a t [ G , A , . . . , A ] = 1. T h e n A i s s o l v a b l e , a n d i t s d e r i v e d l e n g t h i s a t most m - 1 .

m

P r o o f . W e begin by reformulating the result i n such a way that we need not assume t h a t the action is faithful. I n the general case, we cannot hope

134

4.

Commutators

t o prove that A is solvable, b u t we show b y i n d u c t i o n o n m t h a t A ^ ~ ^ C C ( G ) , where we recall t h a t C ( G ) is the kernel of the action. I n the original situation, where A acts faithfully, we have C { G ) = 1, a n d it w i l l follow t h a t A ^ - ) = 1, a n d so A is solvable w i t h derived length at most m - 1, as required.
A A A 1

If m = 1, then [ G , A ] = 1, a n d thus A = C ( A ) a n d there is n o t h i n g more to prove. A s s u m e then, that m > 2, a n d let H = [ G , A ] , so that H admits A a n d [ H , A , . . . , A ] _ i = [ G , A , . . . , A ] = 1. B y the inductive hypothesis applied to the action of A o n H , therefore, we have A ^ " ) C C ( H ) , a n d thus 1 = [ H , A ^ - ^ ] = [ G , A , A ^ ~ \ Since A D A ( " ) , it follows t h a t [ G , A ^ ~ \ A ^ ~ ^ ] = 1, a n d thus also [ A ^ ~ \ G, A ( " ) ] = 1 since interchanging the first two entries i n a m u l t i p l e c o m m u t a t o r of subgroups has no effect. T h e three-subgroups l e m m a now yields
G m m 2 m m m 2 A m 2 m 2 m 2 m 2

[A^ - \A^ - \G} a n d since b y definition,

j ^ m - l )

=l ,

_ ^ ( m - 2 ) ^(m-2)j
m 1

we have [ A ^ ~ \ G \

= 1. T h u s A ^ ' ^ C C ( G ) , as wanted.
A

W h e n m > 2, the step i n the proof of T h e o r e m 4.22 where A is replaced by A ( " ) throws away information, a n d this suggests t h a t perhaps more is true t h a n we proved. I n fact more is true: P . H a l l showed t h a t under the hypotheses of our theorem, A is a c t u a l l y nilpotent, a n d not merely solvable. A l s o , the nilpotence class of A is at most m { m - l ) / 2 . R e c a l l that b y C o r o l l a r y 4.13, the derived length of a nilpotent group is l o g a r i t h m i c a l l y b o u n d e d i n terms of its nilpotence class, a n d therefore it follows b y H a l l ' s theorem t h a t the linear b o u n d o n the derived length of A that we established in T h e o r e m 4.22 is too weak; it is of the "wrong" order of magnitude.
m 2

If m = 2, then A ^ ~ ^ = A ^ = A , a n d we lose n o t h i n g i n the replacement step of the previous proof. I n this case, H a l l ' s result coincides w i t h T h e o r e m 4.22 since a n o n t r i v i a l group is solvable w i t h derived length 1 i f a n d o n l y i f it is abelian, w h i c h is exactly when it is nilpotent w i t h class 1. We list this case as a separate corollary. 4.23. C o r o l l a r y . L e t A a n d G be g r o u p s . Suppose t h a t A acts f a i t h f u l l y o n G v i a a u t o m o r p h i s m s a n d t h a t [ G , A , A] = 1. T h e n A i s a b e l i a n . P r o o f . T h i s is just the case m = 2 of T h e o r e m 4.22.

W i t h o u t using a d d i t i o n a l machinery, it seems to be difficult to establish H a l l ' s upper b o u n d o n the nilpotence class of A . B u t it is not too h a r d to show t h a t A is nilpotent i n the s i t u a t i o n we have been discussing, a n d the proof offers more opportunities for using the three-subgroups l e m m a .

4C

135

4.24. T h e o r e m . L e t A a n d G be finite g r o u p s . Suppose t h a t A acts faithf u l l y v i a a u t o m o r p h i s m s o n G, a n d assume t h a t [ G , A , . . . , A ] 1. T h e n A is n i l p o t e n t .

P r o o f . A s before, we drop the assumption t h a t A acts faithfully. W e prove by i n d u c t i o n on \ G \ that i n general, the last t e r m A of the lower central series of A acts t r i v i a l l y on G. Once this is established, we see t h a t if the action of A is faithful, then A = 1, and hence A is nilpotent. Since we can assume that G > 1 and we k n o w that [ G , A , . . . , A ] = 1, it follows t h a t [ G , A ] < G. A l s o , [ G , A ] admits the action of A and [ [ G , A ] , A , . . . , A ] = 1. B y the inductive hypothesis, therefore, A acts t r i v i a l l y on [ G , A ] , and hence [ G , A , A] = 1, or equivalently, [ A , G, A } = 1. O u r goal w i l l be to show that [ G , A,A] = 1 since, i f we can accomplish that, the three-subgroups l e m m a w i l l y i e l d [A,A,G] = 1. B u t [A,A] = A because A is the last t e r m of the lower central series of A , a n d it w i l l follow that [A, G] = 1. T h e proof w i l l then be complete.
0 0

O u r strategy for showing that [G,A,A] = 1 is to find a n o n t r i v i a l n o r m a l subgroup C of G on w h i c h A acts t r i v i a l l y . Once we have found C , we can a p p l y the inductive hypothesis to the i n d u c e d action of A on G = G / C . (This is v a l i d since we have [ G , A , . . . , A ] = [ G , A , . . . , A ] = 1.) B y the i n d u c t i v e hypothesis, 1 = [G,A] = [G,A], a n d so [G,A] C C. Since we are assuming that A acts t r i v i a l l y on C, this yields [ G , A,A} = 1, w h i c h is what we want. We can assume t h a t [ G , > 1, or else there is n o t h i n g to prove. Take C = C ( A ) , so t h a t A certainly acts t r i v i a l l y o n C, as required. T o see t h a t C > 1, observe that [ [ G , A],A, . . . , A ] = 1 , a n d thus if we start w i t h the n o n t r i v i a l group [ G , A } and repeatedly compute commutators w i t h A , we get a sequence of subgroups of [ G , A] t e r m i n a t i n g at the identity. T h e last nonidentity t e r m i n this sequence, therefore, is centralized b y A , and this shows t h a t C > 1, as required.
l G i A o o ] 0 0

To complete the proof now, it suffices to show t h a t C < G, and our argument for this involves two further applications of the three-subgroups l e m m a . F i r s t , we argue that [G,A] centralizes [ G , A } . W e have already mentioned that A acts t r i v i a l l y on [ G , A ] , so t h a t [[G,A\,A} = 1, and thus we have [[G,A],A,G] = 1. A l s o , [ G , A ] < G, a n d thus [ G , [ G , A } } C [G, A ] , and we have [G, [G,A],A] C [[G, A ] , A] = 1. N o w we apply the three-subgroups l e m m a to deduce that [A, G , [G, A}] = 1, and thus [G, A] centralizes [A, G] = [G, A ) , as claimed. I n p a r t i c u l a r , since C C [G, A } , we have [ G , A , C ] = 1. A l s o , of course, [ A , C , G ] = 1 since A centralizes C. B y the three-subgroups lemma, once again, we have [G, G , A ] = 1, and so A centralizes [G, G ] .
0 0 0 0

136

4.

Commutators

R e c a l l now t h a t C C [G,A] < G, and thus [ C , G ] C [G,A]. W e just showed, however, t h a t A centralizes [ C , G ] , and thus [ C , G ] C C =o](A) = C , where the equality holds by the definition of C. W e have now shown t h a t [ C , G] C C , or equivalently, that C < G, as wanted. T h e proof is now complete.
[ G i X

Let us continue to assume that [ G , A , . . . , A ] = 1 for some unspecified number of copies of A . W e have seen that i f the action is faithful, then A must be nilpotent. It is impossible, however, to draw any conclusions about the structure of G since, for example, it could be that A is t r i v i a l . B u t although we can say n o t h i n g about G itself, we shall see that the subgroup [ G , A ] is under control. W e begin w i t h a special case. 4 .25. Lemma. L e t A a n d G be g r o u p s . Suppose t h a t A acts o n Gv i a

a u t o m o r p h i s m s , a n d assume

t h a t [ G , A , A ] = 1. T h e n [ G , A ] i s a b e l i a n .

P r o o f . Since [ [ G , A } , A ] = 1, we certainly have [ [ G , A ] , A , G ] = 1. A l s o , [ G , A] < G, and thus [ G , [ G , A ] , A ] C [[G, A ] , A] = 1. B y the three-subgroups l e m m a , we have 1 = [A,G, [G,A\] = [[G,A],[G,A]] = [G, A}', and so [ G , A ] is abelian. 4 .26. T h e o r e m . via automorphisms o na finite

L e t A be a p - g r o u p t h a t acts

g r o u p G, a n d suppose

t h a t [ G , A , . . . , A ] = 1. T h e n [ G , A ) i s a p - g r o u p .

P r o o f . W e proceed by i n d u c t i o n on \ G \ . W e can certainly assume that G > 1, and thus since [ G , A , . . . , A ] = 1, it follows that [ G , A ] < G. For convenience, write N = [ G , A ] , and observe that N < G and that N admits the action of A . Since N < G , the inductive hypothesis applies i n N , and we deduce t h a t [ N , A ] is a p-group. A l s o , since [TV, A ] < N , we have [AT, A ] C U, where U = O ( N ) is characteristic i n N < G, a n d thus U < G .
p

Now A

consider the induced action of A on G = G / U .

Since [TV, A ] C U,

we see t h a t A acts t r i v i a l l y on N . A l s o , [G, A ] = [ G , A \ = N , and thus acts t r i v i a l l y on G/N. W e Jiave J G , A , A] = [ N , A ] = 1, and thus by
p

L e m m a 4.25, we conclude t h a t N = [ G , A ] is abelian. B u t O ( N ) a n d since N is abelian, it follows that N is a p'-group.

is t r i v i a l ,

Since A is a p-group and N is a n o r m a l p'-subgroup of G, we can apply C o r o l l a r y 3.28 to the action of A on G. T h a t corollary tells us that "fixed points come from fixed points" ,_and_ since N is contained i n the fixed-point subgroup and every element of G / N is A - f i x e d , we conclude t h a t the whole group G consists of fixed points. T h e n 1 = [ G , A ] = [ G , A ] , and hence [ G , A ] C U. Since U is a p-group, we are done.

4 D

137

In the s i t u a t i o n of T h e o r e m 4.26, we can conclude t h a t [G, A] is nilpotent even w i t h o u t assuming that A is a p-group. 4.27. T h e o r e m . L e t A a n d G be finite g r o u p s . Suppose t h a t A acts v i a a u t o m o r p h i s m s o n G, a n d assume t h a t [ G , A , . . . , A ] = 1. T h e n [ G , A ] i s nilpotent. P r o o f . W e proceed by i n d u c t i o n on \ A \ . ( O f course, i f A is t r i v i a l t h e n [ G , A ] = 1 and there is n o t h i n g to prove.) Suppose that B is an a r b i t r a r y proper subgroup of A . T h e n [ G , B , . . . , B ] C [ G , A , . . . , A ] = 1, and so by the i n d u c t i v e hypothesis, [ G , B ] is nilpotent. Since [G, B ] < G , we have [G, B ) C F ( G ) , the F i t t i n g subgroup, and thus B acts t r i v i a l l y on G / F ( G ) . We now k n o w that every proper subgroup of A is contained i n the kernel of the i n d u c e d action of A on G / F ( G ) , and so i f A is generated by its proper subgroups, then this kernel must be a l l of A . I n this case, [ G , A ] C F ( G ) , and hence [G, A ] is nilpotent, as wanted. W e can assume, therefore, that A is not generated by its proper subgroups. E v e r y finite group, however, is generated by its Sylow subgroups, a n d thus some Sylow subgroup of A is not proper. I n other words, A is a p-group for some p r i m e p , and hence [G, A ] is a p-group by T h e o r e m 4.26. In particular, [G, A ] is nilpotent.

Problems 4C 4C.1. L e t A act o n G v i a automorphisms, a n d suppose t h a t 1 = H

0

C H i C C H

= G

is a chain of A - i n v a r i a n t subgroups such t h a t each right coset of H ^ i n H i is A - i n v a r i a n t for a l l i w i t h 0 < i < m. I n the older literature, this s i t u a t i o n was described by saying that A stabilizes the chain { H i } . If A stabilizes a chain of subgroups i n G , show that it stabilizes a c h a i n of s u b n o r m a l subgroups. 4 C . 2 . A s s u m e that A acts faithfully on G v i a automorphisms, and assume t h a t A stabilizes a chain { H i | 0 < i < m } of n o r m a l subgroups of G . Show t h a t A is nilpotent w i t h class at most m - 1. 4 C . 3 . L e t A act v i a automorphisms on G , and suppose t h a t N < G and t h a t A acts t r i v i a l l y on N . Show that N centralizes [G, A ] , 4D We continue our study of commutators that arise w h e n A acts v i a automorphisms on G , but now we focus on coprime actions, where ( | G | , \ A \ ) = 1.

138

4.

Commutators

The following u l t i m a t e l y relies on G l a u b e r m a n ' s l e m m a , and t h a t is the reason for the solvability assumption, which, as usual, is not really necessary i f one is w i l l i n g to appeal to the F e i t - T h o m p s o n theorem. 4.28. L e m m a . L e t A a n d G be finite g r o u p s . L e t A a c t v i a a u t o m o r p h i s m s o n G, a n d suppose t h a t ( \ G \ , \ A \ ) = 1 a n d t h a t one of A o r G i s s o l v a b l e . ThenG = C (A)[G,A\.
G

P r o o f . W r i t e G = G / [ G , A ] . Since fixed points come from fixed points i n coprime actions ( C o r o l l a r y 3.28) we have C ( A )
U

= C (A).
G

B u t A acts

t r i v i a l l y on G / [ G , A ] , and so the left side of this equation is the whole group G. Thus C (A)[G,A]
G

= C ( A ) = G,
G

and

it follows t h a t [ G , A ] C ( A ) = G by the correspondence theorem.

G

To demystify this proof somewhat, we observe that what is going on here is the following. E a c h coset of [ G , A] i n G is A - i n v a r i a n t , and so by a fairly standard G l a u b e r m a n ' s l e m m a argument, each such coset contains an A - f i x e d element. The following result is used frequently.

4.29. L e m m a . L e t A a c t v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d suppose t h a t ( \ G \ , \ A \ ) = 1. T h e n [ G , A , A ] = [ G , A \ . P r o o f . O f course [ G , A , A] C [ G , A ] , and so it suffices to prove the reverse containment. F o r this purpose, we show that [g, a] G [ G , A , A ] for a l l g G G and a G A . T h e result w i l l then follow since [ G , A ] is generated by these commutators [ g , a ] . Suppose first t h a t A is solvable. L e m m a 4.28 yields G = C ( A ) [ G , A ] , and so i f g G G, we can w r i t e g = cx, where c G C ( A ) and x G [ G , A } . N o w let a G A , and observe t h a t [c, a] = 1 . T h e n
G G

[g, a] = [ c x , a] = [c, a ] [ x , a] = [ x , a] G [ G , A , A] , as wanted. T h i s proves t h a t the result is true i f A is solvable. Now i n the general case, again let g G G and a G A . T h e n [g, a] G [ G , (a)} = [ G , ( a ) , (a)} C [ G , A , A] since the cyclic group ( a ) is certainly solvable. T h i s completes the proof.

We can a p p l y this to the situation that we studied i n the previous section. 4.30. C o r o l l a r y . L e t A a c t f a i t h f u l l y v i a a u t o m o r p h i s m s o n G, w h e r e A a n d G a r e finite g r o u p s , a n d assume t h a t [ G , A , . . . , A ] = 1. T h e n every p r i m e d i v i s o r of \ A \ a l s o d i v i d e s \ G \ .

4D

139

P r o o f . L e t P S y l ( A ) where p is a p r i m e t h a t does n o t d i v i d e | G | . T h e n [ G , P] = [ G , P , . . . , P ] = 1, where the first equality holds by repeated a p p l i c a t i o n of L e m m a 4.29. Since the action of A is faithful and P acts trivially, we deduce that P = 1, and thus p does not d i v i d e \ A \ .
p

N e x t , we prove the so called P x Q l e m m a of T h o m p s o n , and we show how it is relevant to the study of local subgroups. 4.31. T h e o r e m ( T h o m p s o n ) . L e t A be a finite g r o u p t h a t acts v i a a u t o m o r p h i s m s o n a p - g r o u p G, a n d suppose that A = P x Q is a n internal d i r e c t p r o d u c t of a p - g r o u p P a n d a p - g r o u p Q. Suppose that Q fixes every element of G t h a t P fixes. Then Q acts t r i v i a l l y o n G.
1

Before we begin the proof of T h o m p s o n ' s theorem, we establish some easy facts about p-groups acting on p-groups. 4.32. L e m m a . L e t P be a p - g r o u p t h a t acts v i a a u t o m o r p h i s m s t r i v i a l p - g r o u p G. Then [ G , P ] < G a n d C ( P ) > 1.
G

o na non-

P r o o f . T h e semidirect product r = G x P is a p-group, and hence it is nilpo tent. It follows that some repeated c o m m u t a t o r of the form [T, T , . . . , T] is t r i v i a l , a n d thus we have [ G , P , . . . , P ] = 1. Successive c o m m u t a t i o n w i t h P , therefore, yields a descending sequence of subgroups s t a r t i n g w i t h G > 1 and ending w i t h 1. T h u s [ G , P ] < G, as wanted. A l s o , i f H is the last nonidentity group i n this sequence, then [ H , P ] = 1, and so 1< H C C { P ) .
G

A s s u m i n g t h a t the p-group G is n o n t r i v i a l i n the s i t u a t i o n of the P x Q l e m m a , it follows by L e m m a 4.32 that C ( P ) > 1. B y hypothesis, however, C ( P ) C C ( Q ) , and therefore Q has at least some n o n t r i v i a l fixed points in G. W h a t we need, however, is that Q fixes every element of G, and the proof of this, while not hard, is a little more subtle; it gives us another o p p o r t u n i t y to use the three-subgroups l e m m a .
G G G

P r o o f of T h e o r e m 4.31. W e can assume t h a t G > 1, and we proceed by i n d u c t i o n on | G | . W e have [G, P] < G by L e m m a 4.32. A l s o , [ G , P ] is A invariant since b o t h G and P are n o r m a l i z e d by A i n the semidirect p r o d u c t G x > A . T h e hypothesis that Q centralizes every element of G centralized by P is inherited by the a c t i o n of A on [G, P ] , and so the i n d u c t i v e hypothesis applies to this action. W e conclude that Q acts t r i v i a l l y on [G, P ] , and hence [ G , P , Q ] = 1. A l s o , [ P , Q , G ] = 1 since P and Q centralize each other i n A . It follows by the three-subgroups l e m m a that [ Q , G , P ] = 1, and thus P acts t r i v i a l l y on [ Q , G ] = [ G , Q ] . B y hypothesis, Q acts t r i v i a l l y on this subgroup too, and so [G, Q , Q ] = 1. F i n a l l y , since Q is a p'-group and G is

140

4.

Commutators

a p-group, we have [G, Q, Q] = [ G , Q] by L e m m a 4.29, and so [G, Q] = 1, as required. A s an a p p l i c a t i o n of the P x Q l e m m a , we prove the following "local to global" result for p-solvable groups. (Recall from Section 2 C t h a t i f p is a prime, a subgroup H C G is said to be p-local i n G i f H = N ( P ) , for some n o n t r i v i a l p-subgroup P of G.)
G

4.33. T h e o r e m . L e t G be a finite p - s o l v a b l e g r o u p f o r some O , ( H ) C C y (G) / o r every p - l o c a l s u b g r o u p H C G .

p

prime p.

Then

P r o o f . L e t Q = C y ( P ) , where i f is p-local i n G. F i r s t , we consider the case where C y ( G ) = 1, and where our goal, therefore, is to show that Q = l . Since H is p-local, we can write H = N ( P ) for some p-subgroup P C G , a n d we observe t h a t b o t h P and Q are n o r m a l i n P . Since P is a p-group and Q is a p'-group, we have P n Q = 1, and thus P Q is the direct product of P a n d Q .
G

Now let U = O p ( G ) , so that U is a p-group, and observe that P Q acts on U b y conjugation since U < G. W e argue that Q fixes every element of U that is fixed by P , or equivalently, that Q centralizes Q / ( P ) . T o see this, observe that C { P ) C P n N ( P ) = P n P . B u t P n P is a n o r m a l p-subgroup of H , a n d Q is a n o r m a l p'-subgroup of P , and so Q centralizes U n P , a n d hence it centralizes C [ / ( P ) , as wanted. W e conclude by the P x Q l e m m a t h a t Q C C ( P ) . B u t G is p-solvable and we are assuming that C y ( G ) = 1, a n d thus U contains its own centralizer i n G . (This is L e m m a 1.2.3 of H a l l and H i g m a n , w h i c h is our L e m m a 3.21.) W e now have Q Q C ( P ) C U, and since Q is a p'-group and U is a p-group, we deduce that Q = 1, as required i n this case.
V G G G

In the general situation, let TV = C y ( G ) and w r i t e G = G/N, Cy(G) = 1.

so that

If H C G is p-local, then since \ N \ is not divisible by p,

L e m m a 2.17 guarantees that H is p-local i n G , and thus by the first part of the proof, Op/(IF) = 1. N o w Q is a n o r m a l p'-subgroup of P , and so Q Q O / ( P ) = 1. F i n a l l y , since Q = 1, we have Q C N , and the proof is
p

complete.

We return now to the s i t u a t i o n considered at the beginning of this sect i o n . W e h a d A acting c o p r i m e l y on G v i a automorphisms, and either G or A is solvable, and we showed i n L e m m a 4.28 that G = C ( A ) [ G , A ] . A s the following result of H . F i t t i n g shows, more is true i f G is abelian.
G

4.34. T h e o r e m ( F i t t i n g ) . L e t A a c t v i a a u t o m o r p h i s m s o n a n a b e l i a n g r o u p G, a n d assume t h a t A a n d G a r e finite a n d t h a t ( | G | , \ A \ ) = 1. T h e n G = C { A ) x [G, A ] .

G

4D

141

Proof. that G and so and so Let

A s G is abelian, b o t h factors are n o r m a l , a n d so it suffices to show = C ( A ) [ G , A ] and t h a t C ( A ) n [ G , A ] = 1. O f course, G is solvable, we can appeal to L e m m a 4.28 to deduce the first of these two facts, it suffices to prove the second.
G G

9 : G -> G be the m a p denned by 0(g) = H g

beA
b

and note t h a t since G is abelian, the order of the factors i n this p r o d u c t is irrelevant, and hence 9 is well defined. A l s o , i f x , y e G , then ( x y ) = x y for a l l b e A , and so again using the fact t h a t G is abelian, it is easy to see t h a t 9 ( x y ) = 6 ( x ) 9 { y ) , a n d thus 9 is a h o m o m o r p h i s m .
b b b

If g C ( A ) , then g = g for a l l b e A , a n d thus 9 ( g ) = g \ \ . A l s o , since \ A \ is coprime to | G | , we see t h a t g \ \ ^ 1 i f g + 1, a n d thus C ( A ) n k e r ( 9 ) = 1. T o complete the proof, therefore, it is enough to show t h a t [G, A] C ker(0), and since [ G , A] is generated by commutators of the form [g, a ] , w i t h g e G and a e A , it suffices to prove t h a t each of these commutators lies i n ker(0).
G A G

If g e G a n d a e A , then

o(g )
a

= X[g
beA

a b

= e{g),

where the second equality holds because a b runs over A as 6 runs over A . It follows t h a t e([g, a}) = 9 ( g - g ) = 9 { g - ) 9 { g ) = 9 ( g ) - 9 ( g )
1 a l a 1

= 1.

T h i s completes the proof.

A n easy a p p l i c a t i o n of F i t t i n g ' s theorem is the following. 4.35. C o r o l l a r y . L e t A a c t o n G v i a a u t o m o r p h i s m s , w h e r e G i s a n a b e l i a n p - g r o u p , a n d assume t h a t A i s a finite p ' - g r o u p . If A fixes every element of o r d e r p i n G, t h e n A acts t r i v i a l l y o n G. P r o o f . B y F i t t i n g ' s theorem, G = C { A ) x [ G , A ] , a n d by hypothesis, C ( A ) contains every element of order p i n G . Since C ( A ) and [G, A] have o n l y the identity i n c o m m o n , it follows t h a t [ G , A ] contains no elements of order p . B u t [ G , A] is a p-group a n d every n o n t r i v i a l p-group contains elements of order p. W e conclude t h a t [ G , A ] = 1, and the proof is complete.
G G G

Surprisingly, the hypothesis t h a t G is abelian i n C o r o l l a r y 4.35 is unnecessary if p > 2.

142

4.

Commutators

4.36. T h e o r e m . L e t A a c t o n G v i a a u t o m o r p h i s m s , w h e r e G i s a p - g r o u p w i t h p > 2 , a n d assume t h a t A i s a finite p ' - g r o u p . If A fixes every element of o r d e r p i n G, t h e n A acts t r i v i a l l y o n G. The idea of the proof of T h e o r e m 4.36 is quite simple. W e show t h a t if G is a m i n i m a l counterexample, then there is an abelian counterexample of the same order, and this contradicts C o r o l l a r y 4.35. It is straightforward to show t h a t a m i n i m a l counterexample w o u l d have nilpotence class at most 2 (and t h a t does not require that p > 2). B u t getting from class 2 to abelian seems to require a new idea: a remarkable trick due to R . Baer. 4.37. L e m m a (Baer t r i c k ) . L e t G be a finite n i l p o t e n t g r o u p of o d d o r d e r a n d n i l p o t e n c e class a t most 2. T h e n t h e r e exists a naddition operation x + y d e f i n e d f o r elements x,y G G such t h a t G i s a n a b e l i a n g r o u p w i t h r e s p e c t to addition. Also, the f o l l o w i n g hold. (a) If xy = yx f o r x,y <EG, then x + y = - x y . element o f G i s equal to its m u l t i p l i c a t i v e

(b) T h e a d d i t i v e o r d e r of each order. (c) E v e r y a u t o m o r p h i s m g r o u p (G,+).

of G i s a l s o a n a u t o m o r p h i s m

of t h e

additive

Before we begin the proof of Baer's l e m m a , we introduce a bit of suggestive n o t a t i o n . If m is an integer and G is a group of order relatively prime to m, then the m a p x i-> x has an inverse, namely the m a p x ^ x , where n is chosen so t h a t r a n = 1 m o d \ G \ . G i v e n x G G, therefore, there is a unique element y G G such that y = x , a n d i n fact, y is a power of x . In the case where | G | is o d d a n d m = 2, we write y = ^/x~ to denote the unique element w i t h square x , and we observe that if H C G is any subgroup, then V h e H for every element he H . A l s o , it is easy to see that V o F = ( v ^ ) " and that if x a n d y commute, then y / x y = y / x ^ y .
m n m 1 1

P r o o f of L e m m a 4.37. for a l l x,y alent to y - ' x ^ y x y/[y,x], is clearly true. =

Let x + y = x y ^ ^ x ] .
1

T o show t h a t x + y = y + x = x y ^ / \ y ~ r ] . T h i s is equiv-

e G, we must check that y x y ^ ~ y ]

/ravrar -

But

(^iSl)"

= v l ^ / F =
2

and so the identity we must establish is [ y , x ] = ( \ / { y , x \ ) , w h i c h

Observe that if x a n d y commute, then [ y , x ] = 1, a n d since = 1, it follows t h a t x + y = xy, a n d this establishes (a). In p a r t i c u l a r , since 1 and x commute, we have 1 + x = l x = x for a l l x G G, and thus 1 is an additive identity. A l s o , x a n d x ~ commute, and hence x + x ~ = xx~ = 1, and thus x ~ is the additive inverse of x .
l l l l

4D

143

We

have
x

+ ( + ) =
y z

+ yzy/^y]

x y z ^ y \ s ) [ y z ^ y \ , x \ .

Since we are assuming that G has class at most 2, a l l commutators i n G are central, a n d thus the m a p [ - , x ] is a h o m o m o r p h i s m . It follows that [ y z ^ [ z ~ y ] , x] = [y, x] [ z , x] [ y / & y ] , x] , a n d we observe that the t h i r d factor is t r i v i a l since [ z , y] Z ( G ) , and hence y / [ z ~ y ] Z ( G ) . W e now have x + (y + z) = x y z y f [ z ~ y ~ W { y , x } [ z , x ] = xyzy/[z,y][y,x][z,x], [y,x][z,x].

where the second equality holds because [ z , y ] , commutes w i t h O n the other hand, (x + y ) + z = x y ^ x ] + z = xy^x\zyJ[z,xy^/[yTx\}

a n d this can be simplified i n a similar manner. (Here, we use the fact t h a t [ z , } is a homomorphism.) W e o b t a i n (x + y ) + z = xyzy/[y,x][z,x][z,y],

w h i c h is equal to the result of our previous c o m p u t a t i o n since the c o m m u tators commute. T h i s shows t h a t a d d i t i o n on G is associative, a n d hence (G, +) is a group. We have already proved (a). For (b), we use the customary notation for a d d i t i v e l y w r i t t e n groups, and we w r i t e nx to denote the sum of n copies of x , where n is a positive integer a n d x <E G. W e argue by i n d u c t i o n on n t h a t nx = x , w h i c h is certainly v a l i d for n = 1. F o r n > 1, we have
n

nx

= x + ( n- l)x = x + x ~

= xx ~

= x

where the second equality holds by the inductive hypothesis a n d the t h i r d holds by (a), since x and x ~ commute i n G . It follows t h a t nx = 1 if a n d o n l y if x = 1, and since 1 is the additive identity (as well as the m u l t i p l i c a t i v e identity), it follows that the additive a n d m u l t i p l i c a t i v e orders of x are equal.
n l n

F i n a l l y , the a d d i t i o n i n G is uniquely determined by the m u l t i p l i c a t i o n , a n d thus any p e r m u t a t i o n of the elements of G t h a t preserves the m u l t i p l i cation w i l l also preserve the a d d i t i o n . A s s e r t i o n (c) follows a n d the proof is complete. P r o o f o f T h e o r e m 4 . 3 6 . W e can assume t h a t G > 1, a n d we proceed by i n d u c t i o n on G . If H < G and H admits the a c t i o n of A , then since every element of order p i n H is fixed by A , the i n d u c t i v e hypothesis guarantees that A acts t r i v i a l l y on H , and hence [ H , A] = 1. I n particular, if [G, A ] < G, then since [G, A ] admits A , we have 1 = [G, A , A ] = [G, A ] , a n d there is

144

4.

Commutators

n o t h i n g further to prove i n this case. (The second equality holds, of course, by L e m m a 4.29, w h i c h applies since = 1.) The derived subgroup G' is characteristic i n G, and hence it admits A . A l s o , the p-group G is certainly solvable, a n d thus G' < G. W e deduce that [ G ' , A] = 1, a n d so [<?', A , G] = 1. A l s o , since G' < G, we have [ G , G'\ C G', and thus [ G , G ' , A ] C [ G ' , A ] = 1. T h e three-subgroups l e m m a now yields
[ A , G, G'j = 1. B u t [ A , G ] = [ G , A ] = G, a n d so [ G , G'] = 1 and G' CZ(G).

In other words, the nilpotence class of G is at most 2, a n d we can a p p l y the B a e r trick to construct the abelian group structure ( G , +) on the u n d e r l y i n g set of G. Now an element a G A induces an a u t o m o r p h i s m of G, and we conclude by L e m m a 4.37(c) that this p e r m u t a t i o n of the elements of G is also an a u t o m o r p h i s m of ( G , +). In other words, A acts v i a automorphisms on the abelian group ( G , + ) . B y L e m m a 4.37(b), the elements of order p i n ( G , + ) are e x a c t l y the elements of order p i n G, and by hypothesis, they are a l l fixed by A . B y C o r o l l a r y 4.35, therefore, A acts t r i v i a l l y on ( G , + ) , and hence its action on G is t r i v i a l . A similar argument allows us to prove a stronger version of the T h o m p son P x Q l e m m a i n the case where p is o d d . R e c a l l that i n T h e o r e m 4.31 we h a d a group A = P Q acting on a p-group G, where P is a p-subgroup and Q is a p'-subgroup of A , and b o t h are n o r m a l i n A . ( A n d we assumed t h a t C ( P ) C C ( Q ) . ) W e now drop the assumption that P < A .
G G

4.38. Theorem. Let A p - g r o u p G, w h e r e p > 2 , is a p - g r o u p a n d t h a t Q every element of G fixed

be a finite g r o u p t h a t acts v i a a u t o m o r p h i s m s o n a a n d l e t P a n d Q be s u b g r o u p s of A . Suppose that P < A a n d Q i s a p ' - g r o u p . Assume also that Q fixes by P . T h e n Q acts t r i v i a l l y o n G.

P r o o f . F i r s t , observe t h a t b o t h G and Q are n o r m a l i z e d by A i n the semidirect product G x > A . Therefore [ G , Q] is A - i n v a r i a n t , and it admits the action of A . Now assume t h a t G is abelian, so that by F i t t i n g ' s theorem, G = C { Q ) x [ G , Q ] . If [ G , Q ] > 1, then by L e m m a 4.32 a p p l i e d to the act i o n of P on [ G , Q ] , we have C ( P ) > I . T h u s P has n o n t r i v i a l fixed points i n [ G , Q ] , and by hypothesis, these P - f i x e d points are also Q-fixed points. T h i s is a c o n t r a d i c t i o n , however, since [ G , Q ] n C ( Q ) = 1, and it follows t h a t [ G , Q] = 1 i n this case, as wanted.
G [ G i 0 ] G

Now for the general case, we can assume t h a t G > 1. W e proceed by i n d u c t i o n on G, observing t h a t if P < G and H admits the action of A , then the inductive hypothesis yields that [ H , Q] = 1. In particular, since [ G , Q] admits the action of A , we see that if [ G , Q] < G, then 1 =

Problems

4D

145

[G, Q, Q] = [ G , Q ] , a n d we are done. W e c a n assume, therefore, t h a t [ G , Q] = G. N o w G' < G admits the a c t i o n of A , a n d hence [ G ' , Q] = 1, a n d we have [ G ' , Q , G ] = 1 a n d [ G , G ' , Q ] C [ G ' , Q ] = 1. T h u s [ Q , G , G ' \ = 1, a n d since [ Q , G ] = [ G , Q ] = G, we conclude t h a t G' C Z ( G ) . T h u s G has nilpotence class at most 2, a n d we can a p p l y the B a e r t r i c k . Now and A acts o n the abelian group ( G , +) a n d every element fixed b y P

is fixed by Q. B u t we have already proved the t h e o r e m for a b e l i a n groups, so we deduce that Q acts t r i v i a l l y o n ( G , +), a n d hence also o n G .

Problems

4D

4 D . 1 . L e t A act v i a automorphisms o n G , a n d assume t h a t N < G admits A a n d t h a t N D C ( 7 V ) . A s s u m e that (|AT|,

G

= 1.

(a) If A acts t r i v i a l l y on N , show t h a t its a c t i o n o n G is t r i v i a l . (b) If A acts faithfully o n G , show t h a t its a c t i o n o n N is faithful. H i n t . F o r (a) show t h a t [G, A ] C N . C o n s i d e r C ( N ) ,
r G

where T = G x A A l s o , if G is is
G

N o t e . If G is solvable a n d AT = F ( G ) , t h e n N D C ( N ) .
p

p-solvable w i t h C y ( G ) = 1 a n d N = O ( G ) , t h e n TV D C ( 7 V ) . T h u s i n either of these cases, if A acts faithfully a n d c o p r i m e l y o n G , t h e n A acts faithfully o n N . 4 D . 2 . L e t G be nilpotent w i t h class at most 2, a n d assume | G | is o d d . F o l l o w i n g the proof of L e m m a 4.36, construct a n a d d i t i v e structure o n G , and (as u s u a l for a d d i t i v e l y w r i t t e n groups) w r i t e x - y to denote the s u m of x a n d the additive inverse of y . Show t h a t xy - yx = [ x , y } . 4 D . 3 . L e t A act v i a automorphisms o n G , where ( | G | , \ A \ ) = 1, a n d assume t h a t one of A or G is solvable. Suppose t h a t A acts t r i v i a l l y o n every A invariant proper subgroup of G , but that the a c t i o n of A o n G is n o n t r i v i a l . P r o v e a l l of the following. (a) G is a p-group. (b) G' C Z ( G ) . (c) N o A - i n v a r i a n t subgroup H exists w i t h G' < H < G. is elementary abelian.

(d) G I G '

(e) G' is elementary abelian (and possibly t r i v i a l ) . (f) If p > 2, t h e n

4

= 1 for a l l x G G . = 1 for a l l x e G . G/G'.

(g) If p = 2, t h e n x

H i n t . F o r (c), a p p l y F i t t i n g ' s theorem to the a c t i o n of A o n

146

4.

Commutators

4 D . 4 . L e t A act v i a automorphisms on G , where G is a 2-group and A has odd order. Show t h a t i f A fixes every element x G G such that x = 1, then the a c t i o n of A on G is t r i v i a l .
4

4 D . 5 . L e t A be solvable w i t h derived length n , a n d suppose that A acts faithfully v i a automorphisms on an abelian group B , where ( \ B \ , \ A ^ ~ ^ \ ) = 1. Show that G = B x > A has derived length n + 1. Deduce that there exist solvable groups w i t h a r b i t r a r i l y large derived lengths.
n

H i n t . F o r the second part, suppose A has derived length n a n d let C be cyclic of p r i m e order p not d i v i d i n g \ A < - - % Show t h a t if G = G U is the regular w r e a t h p r o d u c t , then G has derived length n + 1.
n

4 D . 6 . L e t # : G -> G be the m a p that appeared i n our proof of F i t ting's theorem (Theorem 4.34.) Show that (in the n o t a t i o n of t h a t theorem) 0 ( G ) = C ( A ) a n d ker(0) = [G, A ] .
G

4 D . 7 . L e t G be p-solvable, and suppose t h a t a Sylow p-subgroup of G is cyclic. L e t K C G be a p'-subgroup, and suppose that p divides | N ( A " ) | . Show that K C C y ( G ) .
G

H i n t . Induct on | G | . If C y ( G ) > 1, a p p l y the i n d u c t i v e hypothesis to G = G/Cy (G). Otherwise, show that G has a n o r m a l Sylow p-subgroup.

Chapter 5

Transfer

5A We seek results t h a t can be used to prove that a group is not simple. In other words, we want to find techniques t h a t can produce n o r m a l subgroups of a group G , or equivalently, h o m o m o r p h i s m s 9 from G into some group H , a n d we want to find hypotheses sufficient to guarantee that k e r ( 9 ) is neither t r i v i a l nor is the whole group G. A s we shall see, the "transfer" h o m o m o r p h i s m , w h i c h we discuss i n this chapter, is a powerful t o o l for p r o v i n g such "good" theorems. W h a t groups H are available to serve as targets for homomorphisms from G ? One possibility, of course, is to take H to be a s y m m e t r i c group S. (Recall that this was how we proved T h e o r e m 1.1 a n d its corollaries.) A n o t h e r useful and well studied s i t u a t i o n is where H is the group of i n vertible n x n matrices over a field F . (Recall t h a t this group is the general linear group, denoted G L ( n , F ) . ) A h o m o m o r p h i s m from G into G L ( n , F ) is called an F - r e p r e s e n t a t i o n of G , a n d representation theory has proven to be a fertile source of good theorems, especially w h e n the field F is taken to be the complex numbers C . W e digress to m e n t i o n t h a t a n F-representation 9 of G determines a function x G - > F , called the character associated w i t h 9, w h i c h is defined by setting x { s ) to be the trace of the m a t r i x 9 ( g ) . Perhaps surprisingly, it turns out t h a t i n the case where F = C , it suffices to study these characters i n order to determine most of the interesting properties of the corresponding representations. I n p a r t i c u l a r , it is not t e r r i b l y difficult to prove that ker(0) = { g e G | x { g ) = x ( l ) } , a n d so characters can be used as proxies for C-representations i n order to prove good theorems. Indeed, it is character theory that p r o v i d e d W . B u r n s i d e w i t h the key to his
n

147

148

5.

Transfer

p Y - t h e o r e m , w h i c h asserts t h a t a group whose order has exactly two prime divisors cannot be simple. We can also choose subgroups of G t o serve as targets for homomorphisms denned o n G. T h a t is approximately what the transfer is: a certain h o m o m o r p h i s m from G into an arbitrary subgroup H . A s we shall see, however, technical considerations necessitate that the target group for the transfer h o m o m o r p h i s m should be abelian, a n d so instead of a subgroup H , we use its c o m m u t a t o r factor group H / H ' as the target. Let G be a n a r b i t r a r y (not necessarily finite) group, a n d let H C G be a subgroup of finite index. I n this section, we define the transfer m a p v : G - H / H ' , a n d we prove t h a t it is a h o m o m o r p h i s m . W e start b y fixing a set of representatives for the right cosets of H i n G. Such a set, called a r i g h t t r a n s v e r s a l for H i n G, is s i m p l y a set of elements of G chosen so t h a t each right coset of H contains exactly one member of the set. If T is a right transversal for H i n G, therefore, the right cosets H t for t G T are distinct, a n d they are a l l of the right cosets of H i n G. W e c a n thus use the elements of T as labels for the right cosets of H , a n d i n particular, \T\ = \ G : H \ . R e c a l l that G acts b y right m u l t i p l i c a t i o n o n the set of right cosets of H , a n d so we c a n write ( H x ) - g = H x g for each right coset H x of H i n G a n d each element g G G. Since the elements of the right transversal T can be viewed as labels for the right cosets of H , the action of G o n the set { H t | t G T } of right cosets uniquely determines a n action of G on T, and we define t - g accordingly. I n particular, ( H t ) - g = H ( t - g ) , or i n other words, if t G T a n d g G G, then t - g is the unique element of T that lies i n the coset ( H t ) - g = H t g . Since our "dot" action of G o n T is really just the action of G o n the right cosets of H , we see, for example, t h a t the stabilizer i n G of an element t G T is exactly the stabilizer of the coset H t , w h i c h we know is H . M o s t importantly, since we know t h a t right m u l t i p l i c a t i o n on right cosets really is an action, it follows t h a t the dot a c t i o n of G o n T is a true action, a n d thus, for example, ( t - x ) - y = t - ( x y ) for t G T a n d x,y G G.
l

Now suppose that t G T a n d g G G. T h e n t g G H t g = H ( t - g ) , a n d hence there is a unique element he H such that t g = h ( t - g ) . Since t g { t - g ) ~ = h , it follows t h a t for a l l elements t G T a n d g G G, we have t g { t - g ) ~ G H . G i v e n g G G, we w o u l d like t o m u l t i p l y a l l of the elements t g i t - g ) - for i e T t o o b t a i n a n element of H depending o n g . There is an obvious difficulty here, however, since i f H is nonabelian, the order i n w h i c h this m u l t i p l i c a t i o n is carried out m a y matter. In order t o resolve fhis issue, we suppose that some specific (but arbitrary) ordering of the elements of T has
l l 1

5A

149

been declared, and we let V : G - > H be the function defined by the formula

teT

where i n c o m p u t i n g the product, the elements t E T are taken i n the specified order. A l t h o u g h this nomenclature is not standard, we refer to any m a p V : G > H constructed i n this way as a pretransfer m a p . T h e r e are usually m a n y different pretransfer maps from G to each subgroup H because, in general, the m a p depends b o t h on the choice of the right transversal T and on the ordering of T. G i v e n a pretransfer m a p V : G - > H , where H C G is an a r b i t r a r y subgroup, we can define a new m a p v : G -> H / H " by composing V w i t h the canonical h o m o m o r p h i s m h -> 7i from H t o H = H / H ' . I n other words ufo) = V ^,
1 7

and we observe t h a t since H / H ' is abelian and each factor t g i t - g ) - i n the product defining V { g ) lies i n I f , it follows t h a t the m a p v is independent of the specific ordering o n T t h a t we used to define V. A s we shall see presently, v is also independent of the choice of the right transversal T, and so for each subgroup H C G, the m a p v is u n a m b i g u o u s l y defined. It is the transfer m a p from G to H / H ' , w h i c h is often inaccurately referred to as the transfer from G to H . (Perhaps we should m e n t i o n t h a t it is customary to use the letter "v" for the transfer because i n G e r m a n , "transfer" is "Verlagerung"). Before we present any theorems or proofs, we introduce some convenient n o t a t i o n . G i v e n two elements x a n d y i n our subgroup H , we can assert t h a t x and y have the same image i n H / H ' b y w r i t i n g H ' x = H ' y or x = y . Unfortunately, if either x or y is some c o m p l i c a t e d expression such as the p r o d u c t defining V { g ) above, the coset n o t a t i o n is somewhat unwieldy, and overbars are even worse. F o r that reason, we introduce the alternative n o t a t i o n x = y to mean the same t h i n g : t h a t x = y . (It w o u l d be more correct to w r i t e x = y m o d H ' , but we w i l l usually suppress explicit m e n t i o n of the m o d u l u s w h e n no confusion is likely to arise.) 5.1. T h e o r e m . L e t G be a g r o u p , a n d suppose t h a t H i s a s u b g r o u p of finite i n d e x . T h e n t h e t r a n s f e r m a p v . G - > H / H ' i s i n d e p e n d e n t of t h e c h o i c e of t h e r i g h t t r a n s v e r s a l used t o d e f i n e i t . P r o o f . L e t S and T be two right transversals for H i n G, a n d observe t h a t for each element t E T, there is a unique element s E S that lies i n the same right coset of H . W e can thus write s = h t for some uniquely determined element h E H , and we have
t t t t

S = {

St

It E T }= {h t \t E T} .
t

150

5.

Transfer

Now let s G S be arbitrary, and let g G G. T h e n s = s for some element i 6 T a n d s - g is the unique element of S i n the coset H s g = H t g = H { t - g ) . It follows t h a t s - g = s . = h . ( t - g ) . W e thus have
t t g t g

sg(s-g)Writing V have
s

= h tg{h . {t-g))- = ht{tg{t-g)- )K;} .

t t g g

and V

to denote pretransfers constructed from S and T , we

Vs(g)

= J]

^(s-g)-

= [J

htitgit-g)- )^.]

Here, the first congruence is not necessarily an equality because we made no assumption about how the a r b i t r a r y orderings on S and T are related. The second congruence is v a l i d because we are w o r k i n g m o d u l o H , and so we are free to rearrange factors t h a t lie i n H . F i n a l l y , t - g runs over a l l of T as t does, a n d thus the first and t h i r d factors on the right cancel (modulo H ' ) , and we conclude that V ( g ) = V { g ) , as required.
1 s T

A l t h o u g h it m a y be reassuring to k n o w that the transfer m a p from G to H is u n a m b i g u o u s l y defined, as we have just shown, the really v i t a l fact about the transfer is that it is a h o m o m o r p h i s m , and the proof of that fact does not rely o n T h e o r e m 5.1. 5.2. T h e o r e m . L e t G be a g r o u p , a n d suppose t h a t H i s a s u b g r o u p of index. Then the transfer m a p v :G - > H / H ' is a h o m o m o r p h i s m . finite

P r o o f . L e t T be a right transversal for H i n G, and let V be a pretransfer associated w i t h T. F o r t G T , we have t - { x y ) = ( t - x ) - y , a n d thus tixy^Hxy))1

= {tx{t-x)- ){{t-x)y{{t-x)-yy ).

Since b o t h factors o n the right lie i n H , we have V(xy) = l[t(xy)(t-(xy))1

= JJ t ^ i - x ) "
= V{x)V{y),

n^xM(i-x)-y)-

where the last congruence holds because t-x runs over a l l of T as i does. It follows t h a t v ( x y ) = v ( x ) v ( y ) , and the proof is complete. A l t h o u g h T h e o r e m 5.2 is not very deep, it enables us to prove the following, w h i c h is our first a p p l i c a t i o n of transfer theory. 5.3. T h e o r e m . Suppose t h a t G i s a finite g r o u p , a n d l e t p be a p r i m e d i v i s o r of \ G ' n Z ( G ) | . T h e n a S y l o w p - s u b g r o u p of G i s n o n a b e l i a n .

5A

151

P r o o f . L e t P G S y l ( G ) , and assume that P is abelian. L e t T be a right transversal for P i n G , and consider the transfer h o m o m o r p h i s m v : G -> P . Since P is abelian, we have u = V, where V is the pretransfer c o m p u t e d using T.
p

Now G' n Z ( G ) is a n o r m a l subgroup of G w i t h order divisible by p, and so its intersection w i t h the Sylow subgroup P is n o n t r i v i a l . W e can thus choose a nonidentity element z G P D G' n Z ( G ) , a n d we compute v { z ) . If t G T , t h e n since z G Z ( G ) , we have P t z = P z t = P t , where the last equality holds because z G P . T h u s i is the element of T i n the coset P i s , and we conclude t h a t t-z = t. T h e n t z { t - z ) ~ = tzt~ = z, where the last equality follows because z is central. W e conclude t h a t v { z ) = V { z ) = z\ \ = z\ \.
l l T G:P

O n the other hand, v is a h o m o m o r p h i s m from G into the abelian group P , a n d hence the derived subgroup G' is contained i n ker(w). Since z G G ' , we have 1 = v ( z ) = z\ \, a n d thus z has order d i v i d i n g the p'-number | G : P | . B u t z G P , and so z has p-power order, and we conclude that z = l . T h i s c o n t r a d i c t i o n shows t h a t P must be nonabelian.
G:P

5.4. C o r o l l a r y . L e t Z C Z ( r ) n T ' , w/iere r i s a finite g r o u p . T h e n a S y l o w p - s u b g r o u p of T / Z i s n o n c y c l i c f o r every p r i m e d i v i s o r p of \ Z \ . P r o o f . L e t P G S y l ( r ) . W e are assuming that p divides \ Z \ and that Z C Z ( r ) n r ' , so it follows by T h e o r e m 5.3 t h a t P is not abelian. Since P n Z C Z ( P ) a n d P is nonabelian, we conclude t h a t P / ( P n Z ) cannot be cyclic. B u t P / ( P n Z ) = P Z / Z G S y l ( r / Z ) , a n d thus T / Z has a noncyclic Sylow p-subgroup.
p p

We digress to provide some context for C o r o l l a r y 5.4. G i v e n a finite group G , consider pairs of groups ( T , Z ) , where Z C Z ( T ) and T / Z = G . (We say t h a t T is a central extension of G i n this situation.) O f course, it is t r i v i a l to find central extensions of G since we could s i m p l y take T = G x Z , where Z is any abelian group. T o prevent this, a n d to make the s i t u a t i o n more interesting, we insist that Z should be contained i n the derived subgroup T' as well as i n the center Z ( r ) . W i t h this a d d i t i o n a l requirement, the p r o b l e m is m u c h more restrictive. Indeed, one can prove that a m o n g all pairs ( T , Z ) , where Z C Z ( T ) n T' a n d T / Z = G , there is a unique (up to isomorphism) largest group Z t h a t can occur as a second component, and t h a t every other possible second component is a h o m o m o r p h i c image of this largest one. T h e largest possible second component of a p a i r (T, Z ) associated w i t h a given group G is called the Schur multiplier of G , and it is denoted M ( G ) . If (T, Z ) is one of our pairs i n w h i c h Z = M ( G ) , then, of course, | T | = \ G \ \ M ( G ) \ , and i n this case, T is said to be a Schur representation g r o u p for G . If the group G is perfect, w h i c h , we recall, means t h a t G' = G ,

152

5.

Transfer

t h e n it has a unique (up to isomorphism) Schur representation group, but in general, a given group G can have more t h a n one i s o m o r p h i s m type of Schur representation group. F o r example, if G is noncyclic of order 4, it is not very h a r d to show t h a t \ M ( G ) \ = 2, and thus b o t h the dihedral group D g and the quaternion group Q are Schur representation groups for G.
8

C o r o l l a r y 5.4 can be restated this way: If p is a prime divisor of \ M ( G ) \ , t h e n a Sylow p-subgroup of G must be noncyclic, and i n particular, it is n o n t r i v i a l . F o r example, the only p r i m e p for w h i c h a Sylow p-subgroup of the alternating group A is noncyclic is p = 2, and it follows that M ( A ) is a 2-group. I n fact, the Schur m u l t i p l i e r of A has order 2. (We cannot resist mentioning that \ M ( A ) \ = 2 for a l l integers n > 4 except n = 6 and n = 7; i n those two cases, the Schur m u l t i p l i e r has order 6. T h i s is another example of what seems to be c o m m o n phenomenon i n finite group theory: general patterns w i t h finite lists of exceptions.)
h 5 b n

Problems 5A 5 A . 1 . Suppose that G is abelian and H C G has index n . Show that the transfer h o m o m o r p h i s m from G to H is the m a p g g .
n

5 A . 2 . Show that the transfer h o m o m o r p h i s m v : G -> G / G ' is just the canonical h o m o m o r p h i s m from G to G / G ' . 5 A . 3 . L e t H C K C G, where \ G : H \ < oo, and suppose that S is a right transversal for H i n K and that T is a right transversal for K i n G. W r i t e ST { s t \ s e S, t e T } . (a) Show t h a t each element of ST is uniquely of the form st w i t h s G S and i G T. Show that S T is a right transversal for H i n G. |G:H| = \ G : K \ \ K : H \ . Let Deduce that

s G -S and t G T, and suppose that g E G. Show t h a t

(si).gr =

S'(i3(i5) )(i-g).

_1

(d) L e t V r : G > i f be a pretransfer c o m p u t e d w i t h respect to T , and let v : A" - Tf/Tf' be the transfer h o m o m o r p h i s m . Show t h a t the composite m a p w : G TJ/TJ' denned by w ^ ) = v { V ( g ) ) is the transfer h o m o m o r p h i s m .
T

5 A . 4 . L e t H C Z ( G ) , and assume t h a t | G : H \ = n < oo. L e t v : G -> H be the transfer h o m o m o r p h i s m . (a) I f h e H , show that = h .
n

5B

153

(b) N o w assume that H is finite a n d t h a t \ H \ is coprime to n . Show that G = H x ker(v). N o t e . T h e fact t h a t H is a direct factor of G i n (b) can also be proved by an appeal to the Schur-Zassenhaus theorem. 5A.5. L e t G be finite, and suppose C is a cyclic n o r m a l subgroup such that

G / C is cyclic. Show that \ M ( G ) \ divides \ C : G'\. H i n t . A p p l y L e m m a 4.6 to a group Y c o n t a i n i n g a subgroup Z such t h a t Y/Z = G and Z C T' n Z ( T ) .
n

5 A . 6 . L e t D be dihedral of order 2 m u l t i p l i e r M ( D ) has order 2.

where n > 2. Show that the Schur

5 A . 7 . L e t B and C be cyclic subgroups of a finite group G, and suppose t h a t B C = G and B n C > 1. A s i n P r o b l e m 5A.5, assume that C G and t h a t | C : G'\ = n . Show that the order of M ( G ) is s t r i c t l y less t h a n n . N o t e . I n p a r t i c u l a r , semidihedral and generalized quaternion groups have t r i v i a l Schur multipliers. 5 A . 8 . L e t A and B be a r b i t r a r y finite groups. (a) Show t h a t \ M ( A x ) | > \M(A)\\M(B)\.

(b) A s s u m i n g that | A | and | B | are coprime, show t h a t M ( A x B ) = M ( A ) x M(). 5B It is often difficult to o b t a i n useful i n f o r m a t i o n about the transfer m a p i f we work d i r e c t l y from the definition. B u t if we appeal to the transfer-evaluation l e m m a , w h i c h we present i n this section, c o m p u t a t i o n s become easier, and transfer becomes more useful as a t o o l for p r o v i n g theorems. We continue to allow our groups to be infinite, but of course, i n order for the transfer from a group G to a subgroup H to be defined, we must assume t h a t the index \ G : H \ is finite. 5 .5. L e m m a (Transfer E v a l u a t i o n ) . L e t G be a g r o u p , a n d l e t H be a s u b g r o u p w i t h finite i n d e x n . L e t V : G - > H be a p r e t r a n s f e r m a p c o n s t r u c t e d u s i n g a r i g h t t r a n s v e r s a l T f o r H i n G, a n d fix a n element g G G. Then t h e r e e x i s t a subset T C T and positive integers n for t e T (depending o n g ) such t h a t t h e f o l l o w i n g h o l d .
0 t 0

(a) t g ^ t -

6 H for all t G T.
0

(b) V { g ) = . n
ten

tg n~ .

154

5.

Transfer

(c)

TH = n . ier
0 t

(d) If g has f i n i t e o r d e r m, t h e n n

i s a d i v i s o r of m f o r a l l i G T .
0

O f course, the p r o d u c t i n assertion (b) is not well defined unless some order i n w h i c h to take the factors is established. T h i s is really irrelevant, however, since we are w o r k i n g m o d u l o H ' and by (a), each factor t g n ~ assertion (b) says that v ( g ) = n *0 * >
nt _1 w h e r e v i s t h e n x

lies i n H . If H happens to be abelian, so that H ' = 1, then of course, transfer map.

P r o o f of L e m m a 5.5. T h e cyclic subgroup (g) C G acts on T v i a our "dot" action, and thus T decomposes into (#)-orbits. L e t T C T be a set of representatives of these orbits (so that each orbit contains exactly one member of T ) a n d let n be the size of the orbit containing t T . O f course, the s u m of the orbit sizes is | T | = n , and this establishes (c). A l s o , (d) is immediate since if o ( g ) = m < oo, then \ ( g ) \ = m, and so the size of each (#)-orbit on T divides m.
0 0 t 0

Now Since

let (!) be a (#)-orbit on T, and let t G T lie i n O, so that \ 0 \ = n .

0 t

O = { t - x \ x (g)} = {t-g* \ i E Z } , we observe t h a t the elements t - g must be distinct for 0 < i < n , and that these are exactly the elements of O. A l s o , it should be clear that t - g = t.
t n t l

R e c a l l that for every element s T , we have s g ( s - g ) H , and the p r o d u c t of these elements (in some order) is equal to V ( g ) . T a k i n g s = t - g \ we have {t-g^git-g^ )for tg
n t 1 1

- sg(s-g)-

a l l integers i. Since t - g

n t

= t, we see that the p r o d u c t of the elements

of this form w i t h 0 < i < n - 1 (taken i n order of increasing i ) is precisely

t

t - \ w h i c h therefore lies i n H . T h i s proves (a), and it shows that except

n t

possibly for the order of the factors, the element t g Since T is the u n i o n of the various orbits and T of these orbits, assertion (b) follows.

t~

is the c o n t r i b u t i o n

to the p r o d u c t denning V ( g ) c o m i n g from the elements of T that lie i n O.

0

is a set of representatives

A s our first a p p l i c a t i o n of the transfer-evaluation l e m m a , we compute the transfer into a central subgroup. 5.6. T h e o r e m . L e t G be a g r o u p , a n d suppose that Z is a central subgroup of f i n i t e i n d e x n . T h e n t h e t r a n s f e r f r o m G t o Z i s t h e m a p g ^ g . T h i s m a p i s therefore a h o m o m o r p h i s m f r o m G into Z .
n

N o t e t h a t since Z is abelian, the transfer to Z really does m a p to Z ; there is no need to work m o d u l o anything. A l s o , observe t h a t T h e o r e m 5.6

5B

155

generalizes b o t h P r o b l e m 5 A . 1 and P r o b l e m 5 A . 4 ( a ) . F i n a l l y , we m e n t i o n t h a t since, i n the s i t u a t i o n of the theorem, G / Z is a group of order n , it is obvious that g G Z . W h a t is not at a l l obvious, however, is t h a t the m a p g i-> g is a h o m o m o r p h i s m , so that { x y ) = x y for a l l x,yG.
n n n n n

P r o o f of T h e o r e m 5.6. Choose a right transversal T for Z i n G and let g G G . Choose a subset T C T and integers n for i G To, as i n the transferevaluation l e m m a (5.5). F o r t G T , therefore, we have t g ^ t ' G Z , and since Z C Z ( G ) , we have
0 4 1 0

t^'f-

= (i^T )' = t ^ i t g ^ t - ^ t = g

n t

B y L e m m a 5.5(b), the product over i G To of these elements is v ( g ) , where v :G Z is the transfer. Since n = n , however, this p r o d u c t is g .
n t

We digress from our study of the transfer to discuss a n a p p l i c a t i o n of T h e o r e m 5.6. 5.7. T h e o r e m (Schur). Suppose T h e n G' i s a finite subgroup. t h a t Z ( G ) has finite i n d e x i n a g r o u p G.

We need a few p r e l i m i n a r y results. F i r s t , we show t h a t i n the s i t u a t i o n of Schur's theorem, there are only finitely m a n y commutators i n G . O f course, this falls far short of the assertion of the theorem since i n general, not every element of G ' is a c o m m u t a t o r . 5.8. L e m m a . L e t T be a r i g h t t r a n s v e r s a l f o r Z ( G ) i n a g r o u p G. Then every c o m m u t a t o r i n G has t h e f o r m [ s , t ] , w h e r e s,t G T . I n p a r t i c u l a r , if | G : Z ( G ) | < oo, t h e n t h e r e a r e o n l y finitely many different commutators. P r o o f . T h e second statement follows from the first since \T\ = \ G : Z ( G ) | . To prove the first assertion, observe t h a t every element of G can be w r i t t e n i n the form xs, where x G Z ( G ) a n d s G T. To complete the proof, therefore, it suffices to show that [ x s , yt] = [ s , t ] , where x , y G Z ( G ) and s,teT. W e have [ x s , y t ] = [ x , y t } [ s , y t ] = [ s , y t ] , where the second equality holds because [ x , y t ] = 1 since x is central. A l s o , [ s , y t ] = [ i / M ] = ([y, s ] % s])" = [i, s } ' = [ s , t ] , where the t h i r d equality holds since y is central.
s - 1 1 1

We

also need the following easy consequence of T h e o r e m 5.6. Then

5.9. C o r o l l a r y . Suppose that G i s a g r o u p a n d t h a t \ G: Z(G)| = n . t h e n t h p o w e r of every c o m m u t a t o r i n G is the identity.

n

P r o o f . T h e m a p g ^ g is a h o m o m o r p h i s m from G into the abelian group Z ( G ) , and thus the c o m m u t a t o r subgroup G' is contained i n its kernel. The final ingredient i n the proof of T h e o r e m 5.7 is the following somewhat more general fact due to A . P . D i e t z m a n n .

156

5.

Transfer

5 . 1 0 . T h e o r e m ( D i e t z m a n n ) . L e t G be a g r o u p , a n d l e t X C G be a finite subset t h a t i s c l o s e d u n d e r c o n j u g a t i o n . Assume that there is a positive i n t e g e r n such t h a t x = 1 f o r a l l x e X . T h e n ( X ) i s a finite s u b g r o u p of G.

n

P r o o f . L e t S be the subset of G consisting of a l l products of finitely m a n y members of X , allowing repeats, and allowing the "empty" p r o d u c t , w h i c h is equal to 1. T h e n S is closed under m u l t i p l i c a t i o n , a n d also, since x = 1 for each element x G X , we see that x ~ = x ' lies i n S. It follows t h a t S is closed under inverses, and hence S = ( X ) . E v e r y element of ( X ) , therefore, is a p r o d u c t of members of X , and we w i l l show t h a t it is never necessary to use more t h a n (n - 1 ) \ X \ factors. Since X is finite, it w i l l follow t h a t ( X ) is finite, as wanted.
n l 7 1 1

Consider an a r b i t r a r y p r o d u c t g = x i x - - - x of TO not necessarily distinct elements of X , and suppose t h a t some p a r t i c u l a r element x X occurs at least k times a m o n g the factors x where k > 1. W e argue that g can be rewritten as a p r o d u c t of m elements of X i n such a way that the first factor is x , and there are still a t o t a l of at least k factors equal to x . T o see w h y this is true, suppose that t is the smallest subscript such that x = x . W e c a n assume t h a t i > 1, and we observe t h a t
2 m u t

X \ X

- \ X

X \ X

- \ X = =

X ( X \ X X ( X
1

{ X 2 )
X

- l )

{x - )
t l

T h e expression on the right is a product of t elements of X because X is closed under conjugation by x , and of course, its first factor is x . A l s o , there are at least k - 1 copies of x among the elements x w i t h t < i < TO, a n d so i f we m u l t i p l y the above expression on the right by x i x , we get a new representation of g as a product of TO elements of X , and these include a t o t a l of at least k copies of x . T h e first factor i n this p r o d u c t is x , as desired.
{ t + m

A s s u m i n g now (as we may) that x = x , suppose t h a t k > 2. W e can repeat the above argument for the product x x x , a n d so we can assume t h a t x = x . C o n t i n u i n g like this, we see that we c a n assume t h a t g is a p r o d u c t of m elements of X such that the first k factors are a l l equal to x .
x 2 3 m 2

N o w fix an a r b i t r a r y element g ( X ) and w r i t e g = x x x x , where the factors 6 X are not necessarily distinct, a n d where m is as s m a l l as possible. S u p p o s i n g t h a t m > ( n - l ) \ X \ , we derive a c o n t r a d i c t i o n . B y the "pigeon-hole" principle, at least one element x X must occur at least n times among the factors x for 1 < i < TO, and by the above reasoning, it is no loss to assume that the first n factors are a l l copies of x . B u t x = 1,
x 2 % m X i % n

5C

157

and

thus g = x

is a p r o d u c t o f m - n elements of X.

T h i s is

the desired c o n t r a d i c t i o n by the assumed m i n i m a l i t y of m .

F i n a l l y , we assemble the pieces to o b t a i n a proof of Schur's theorem. P r o o f of T h e o r e m 5.7. L e t X be the set of a l l c o m m u t a t o r s i n G, a n d observe t h a t \X\ is finite by L e m m a 5.8. x, A l s o , since [x,y]*>
n

= [x,y }

for The

y a n d g i n G, we see t h a t X is closed under conjugation. N o w w r i t e = 1 for a l l xeX.

n = \ G : Z ( G ) | , so t h a t by C o r o l l a r y 5.9, we have x result now follows by T h e o r e m 5.10.

Problems 5B 5 B . 1 . L e t G be finite, a n d suppose t h a t P G S y l ( G ) a n d that g G P has

p

order p . If g G G', but g P ' , show that ^ with i 0 P .

G P ' for some element i G G

H i n t . L e t u be the transfer h o m o m o r p h i s m from G to P / P ' a n d observe t h a t g G ker(u), a n d thus V ( g ) G P ' , where V is a corresponding pretransfer map. A p p l y the transfer-evaluation l e m m a , a n d note t h a t each integer n is either 1 or p .
t

5 B . 2 . I n the s i t u a t i o n of T h e o r e m 5.10, suppose | A | = m. \ ( X ) \ < n . H i n t . Write X 0 < e* < n . = { x

e 2 u e 2 m

Show t h a t

x ,..., x
2 m

m e

a n d show t h a t every element of ( X )

%

has the form ( x i ) * ( x )

( x ) , where the e

are integers satisfying

5 B . 3 . L e t G be an a r b i t r a r y group. Show t h a t a n element x G G lies i n some finite n o r m a l subgroup of G i f a n d o n l y if x has finite order a n d its conjugacy class i n G has finite size. 5C We and H/H' r e t u r n now to finite groups. If we want to use transfer theory to prove C G, we must show t h a t the kernel of the transfer h o m o m o r p h i s m v : G is neither t r i v i a l nor the whole group G. If P is proper i n G, t h e n (Recall T o prove "good" n o n s i m p l i c i t y

t h a t some group G is not simple, we need to choose a subgroup P

automatically, k e r ( v ) > 1 since i n t h a t case, v cannot be injective. t h a t we are assuming that G is finite.)

theorems, therefore, it is sufficient to find conditions t h a t w i l l guarantee t h a t ker(?j) < G, or equivalently that there exists at least one element x G G such t h a t v ( x ) ^ 1.

158

5.

Transfer

G i v e n a n element x G G , it is sometimes easier to compute v ( x ) if x happens to lie i n the transfer target H . O f course, i f we l i m i t our attention to elements of H and we can show that not every element of H lies i n ker(u), this is good enough. (Obviously, i f H % ker(u), then ker (?j) < G, and i n that case, G is not simple.) I n fact, i f H is a Sylow subgroup of G, or more generally, if H is a H a l l subgroup, then the converse of this assertion is true too: if k e r ( v ) < G, then H % ker(v). I n the s i t u a t i o n where H is a H a l l subgroup, therefore, we lose n o t h i n g by c o m p u t i n g the transfer only on elements of H . T h e following easy l e m m a proves somewhat more t h a n this. 5.11. L e m m a . L e t H be a H a l l s u b g r o u p of a f i n i t e g r o u p G, a n d l e t v : G - > H / H ' be t h e t r a n s f e r h o m o m o r p h i s m . T h e n v { H ) = v { G ) , a n d so \ H : H D k e r { v ) \ = |G:ker(u)|. P r o o f . W e k n o w that \ G : ker (7j)| = \ ( G ) \ , and since v { G ) C H / H ' , we deduce that \ G : ker(v)| divides \ H \ . T h e n \ G : ker(v)| is coprime to \ G : H \ , and so k e r ( v ) H = G by L e m m a 3.16. T h e result now follows.
V

A g a i n let H C G, and consider the transfer h o m o m o r p h i s m v : G - > H / H ' . W h e n we a p p l y the transfer-evaluation l e m m a to compute v ( x ) for some element x G G, we are led to consider elements of H of the form t x t ~ for certain elements i e G and positive exponents n . U x e H , w h i c h as we now k n o w is often the case of interest, then also x G H . I n this situation, the elements x and t x t ~ b o t h lie i n H , and they are clearly conjugate in G ( v i a the element i ) , but of course, they need not be conjugate i n H .
n t 1 t n t n t n t l

M o t i v a t e d by this, we consider the intersection of a conjugacy class X of G w i t h a subgroup H , where we assume that X n H is nonempty. Clearly, if x G A n J f and y lies i n the conjugacy class of H c o n t a i n i n g x , then y X D H , a n d this shows t h a t I n i i i s a u n i o n of conjugacy classes of H . W e say t h a t classes Y\ and F of H are fused i n G i f Y and F are contained i n the same class of G . A l s o , if H C AT C G , we say that AT controls G - f u s i o n i n H if every two classes of H t h a t are fused i n G are already fused i n K . Equivalently, AT controls G-fusion i n H if and only i f every pair of G-conjugate elements of H are ^ - c o n j u g a t e . I n particular, G always controls G-fusion i n H , and H controls G-fusion i n itself precisely when no two distinct classes of H are fused i n G . A s we shall see, the n o t i o n of "control of fusion" is v i t a l for understanding transfer, especially transfer to Sylow subgroups, and more generally, to H a l l subgroups.
2 x 2

We

begin our s t u d y of fusion w i t h the following general result.

p

5.12. L e m m a . L e t P G S y l ( G ) , w h e r e G i s a f i n i t e g r o u p . controls G-fusion i n C (P).

G

Then N (P)
G

5C

159

P r o o f . A s s u m i n g t h a t x , y G C { P ) are conjugate i n G , we must show t h a t in fact, these elements are conjugate i n N ( P ) . W r i t e y = x w i t h g G G, a n d observe t h a t since x G C ( P ) , we have y = x G C { P ) = C {P ). T h e n P C C ( y ) , a n d hence P is a S y l o w p-subgroup of C ( y ) . Also y G C ( P ) , and so P C C ( y ) , and hence P is also a Sylow p-subgroup of C ( y ) . W e can now a p p l y the Sylow C - t h e o r e m i n the group C { y ) to the two Sylow p-subgroups P and P , and we deduce t h a t ( P ) = P , for some element c C ( y ) . T h e n #c G N ( P ) and x ^ = y = y , and so x a n d y are conjugate i n N ( P ) , as required.
G 9 G 9 9 9 G G G 9 9 G G G G G G 9 9 c c c G G G

We can now prove B u r n s i d e ' s n o r m a l p-complement theorem. (Recall t h a t i f p is a prime, then a n o r m a l p-complement i n a finite group G is a n o r m a l subgroup w i t h index equal to the order of a Sylow p-subgroup. E q u i v a l e n t s , it is a subgroup N < G such t h a t \ N \ is not divisible by p and \ G : N \ is a power of p. In s t i l l other words, a n o r m a l p-complement is a n o r m a l H a l l p'-subgroup. A group t h a t has a n o r m a l p-complement is said to be p - n i l p o t e n t . ) 5 . 1 3 . T h e o r e m ( B u r n s i d e ) . L e t P G S y l ( G ) , w h e r e G i s a finite a n d suppose P C Z ( N ( P ) ) . Then G has a n o r m a l p - c o m p l e m e n t .
p G G

group,

P r o o f . I n this s i t u a t i o n , P is abelian, so P C C ( P ) . B y L e m m a 5.12, therefore, every two elements of P t h a t are conjugate i n G are conjugate i n N ( P ) . B u t P is central i n N ( P ) , so no two distinct elements of P can be conjugate i n N ( P ) . It follows t h a t no two distinct elements of P can be conjugate i n G .
G G G

Now consider the transfer h o m o m o r p h i s m v : G -> P , and recall t h a t in this case, where P is abelian, there is no d i s t i n c t i o n between the transfer and pretransfer. L e t x G P , and a p p l y the transfer-evaluation l e m m a to compute t h a t v ( x ) = J]
teT
0
0

tx H~ ,

where T is some subset of a right transversal T for P i n G . A l s o , n = \ G : P | , a n d we k n o w that each of the factors t x ^ t ' lies i n P . Since tx n~ and x are G-conjugate elements of P , the first paragraph of the proof guarantees t h a t they are equal, a n d thus x = t x n ~ for a l l t G T . Then
t 1 n l n t n t n l 0

v(x)

=Y[x
ten

| G : F |

and so if v ( x ) = 1, t h e n x has order d i v i d i n g \ G : P | . B u t x G P , and so the order of x also divides | P | . It follows t h a t x = 1, a n d hence P n k e v ( v ) = 1. B y L e m m a 5.11, we have | G : ker(u)| = | P | , a n d so ker (?j) is the desired n o r m a l p-complement i n G .

160

5.

Transfer

5.14. C o r o l l a r y . L e t P e S y l ( G ) , w h e r e G i s a finite g r o u p a n d p i s t h e s m a l l e s t p r i m e d i v i s o r of \ G \ , a n d assume t h a t P i s c y c l i c . T h e n G has a normal p-complement.

p

P r o o f . W e k n o w t h a t N ( P ) / C ( P ) can be i s o m o r p h i c a l l y embedded i n A u t ( P ) , w h i c h , because P is cyclic, has order < p ( \ P \ ) , where <p is E u l e r ' s function. W e can certainly assume that P > 1, a n d so we can w r i t e \ P \ = p " , w i t h n > 0. T h e n < p ( \ P \ ) = p ( p - 1), and since this number has no p r i m e divisor larger t h a n p, it follows that | N ( P ) : C ( P ) \ has no p r i m e divisor larger t h a n p . T h i s index also has no p r i m e divisor smaller t h a n p because no such p r i m e divides \ G \ . F i n a l l y , the p r i m e p itself does not divide | N ( P ) : C ( P ) | since P is abelian, a n d thus the full Sylow p-subgroup P is contained i n C ( P ) . W e conclude that | N ( P ) : C ( P ) | has no p r i m e divisors at a l l , and so this index is 1, and N ( P ) = C ( P ) . In other words, P C Z ( N ( P ) ) , a n d hence by B u r n s i d e ' s theorem, G has a n o r m a l p-complement.
G G n _ 1 G G G G G G G G G G

We m e n t i o n t h a t if some group G has a cyclic Sylow p-subgroup, then the same is true for every subgroup of G and every h o m o m o r p h i c image of G. (For subgroups, this follows b y the Sylow D - t h e o r e m since a Sylow p-subgroup of a subgroup of G is contained i n a Sylow p-subgroup of G. For factor groups, our assertion is a consequence of P r o b l e m I B . 5 , w h i c h implies that a surjective h o m o m o r p h i s m carries Sylow p-subgroups to Sylow p-subgroups.) I n particular, this shows that the hypotheses of the following corollary are inherited by subgroups and factor groups. 5.15. C o r o l l a r y . L e t G be a finite g r o u p , a n d suppose g r o u p s of G a r e c y c l i c . T h e n G i s s o l v a b l e . that all Sylow sub-

P r o o f . W e c a n assume that G > 1, and we proceed by i n d u c t i o n on \ G \ . If p is the smallest p r i m e divisor of \ G \ , then G has a n o r m a l p-complement N by C o r o l l a r y 5.14. A l s o , N < G a n d a l l Sylow subgroups of TV are cyclic, and thus N is solvable b y the i n d u c t i v e hypothesis. Since G / N is a p-group, it is solvable too, and thus G is solvable by L e m m a 3.10(e). In p a r t i c u l a r , it follows t h a t the order of a nonabelian simple group

cannot be a product of distinct p r i m e numbers. A c t u a l l y , m u c h more c a n be said about groups i n w h i c h a l l Sylow subgroups are cyclic. N o t o n l y is such a group solvable, but it is also metacyclic, w h i c h means t h a t it has a cyclic n o r m a l subgroup w i t h a cyclic factor group. T h e following shows t h a t even more is true. 5.16. T h e o r e m . Suppose t h a t a l l S y l o w s u b g r o u p s of t h e finite g r o u p G a r e c y c l i c . T h e n b o t h G' a n d G / G ' a r e c y c l i c , a n d they h a v e c o p r i m e o r d e r s .

5C

161

The

key step here is the following. group

5.17. T h e o r e m . L e t P be a c y c l i c S y l o w p - s u b g r o u p of some finite G. T h e n p d i v i d e s a t most one of t h e n u m b e r s \ G ' \ a n d \ G : G ' \ .

G

P r o o f . L e t N = N ( P ) , and let K C N be a complement for P i n N . (The existence of K is guaranteed by the Schur-Zassenhaus theorem.) B y F i t t i n g ' s theorem (4.34), we have P = [ P , K ] x C ( K ) . If [ P , K ] = 1, then C { K ) = P , a n d i n this case, P is central i n N = P K . B y B u r n s i d e ' s theorem, therefore, G has a n o r m a l p-complement M , a n d since G / M is a p-group, it must be cyclic. T h e n G' C M , a n d so p does not divide \ G ' \ .
P P

If, on the other h a n d , [ P , K ] > 1, then [ P , K ] contains the unique subgroup of order p i n the cyclic group P . It follows t h a t C ( K ) contains no subgroup of order p, and hence C ( K ) = 1. Therefore P = [ P , K ] C G', and so \ G : G'\ is not divisible by p , as required.
P P

We need one further observation: an abelian group i n w h i c h a l l Sylow subgroups are cyclic must itself be cyclic. I n fact, i f we choose a generator x of the Sylow p-subgroup of G for each p r i m e divisor p of \ G \ , it is easy to see t h a t the element T\ is a generator for G.
p X p

P r o o f of T h e o r e m 5.16. T h a t no p r i m e can divide b o t h \ G ' \ and \ G : G'\ follows b y T h e o r e m 5.17. A l s o , G / G ' is a n abelian group w i t h a l l Sylow subgroups cyclic, a n d hence it must be cyclic. T o show t h a t G' is cyclic, we proceed b y i n d u c t i o n on \ G \ . W e can assume, of course, t h a t G > 1, and thus G' < G because G is solvable b y C o r o l l a r y 5.15. T h e i n d u c t i v e hypothesis applied to G' tells us that G" is cyclic, a n d thus A u t ( G " ) is abelian. B u t G / C ( G " ) is isomorphically embedded i n A u t ( G " ) , and thus G / C ( G " ) is abelian, and we have G' C C ( G " ) . N o w G ' / G " is abelian and hence is cyclic, a n d since G" C Z ( G ' ) , it follows t h a t G' is abelian, and hence G ' is cyclic, as required.
G G G

N e x t , we consider what happens if only part of an abelian Sylow psubgroup P of G is central i n N ( P ) . (In the case of B u r n s i d e ' s theorem, where the whole of P is central i n N ( P ) , it is a u t o m a t i c t h a t P is abelian, but i n this more general s i t u a t i o n we must assume t h a t P is abelian.) Since B u r n s i d e ' s theorem (5.13) is an easy consequence of the following, we c o u l d have o m i t t e d the proof of that earlier result, w h i c h is hereby rendered redundant. W e m e n t i o n also t h a t the following c a n also be viewed as a stronger form of T h e o r e m 5.3.
G G

5.18. T h e o r e m . L e t P be a n a b e l i a n S y l o w p - s u b g r o u p of a Then G ' n P n Z ( N ( P ) ) = 1.

G

finite

g r o u p G.

162

5.

Transfer

P r o o f . L e t v : G - > P be the transfer, and l e t i G G ' n F f l Then 1 = (a;) = J T ^ i v 1

Z(N (P)).
G

where the first equality holds since x e G ' C ker(v), and the second follows by the transfer-evaluation l e m m a , where T and the integers n are as usual, and where each factor t x ^ t ' lies i n P . Since x P , b o t h x and ta^r lie i n P , and hence these G-conjugate elements of P C C ( P ) are conjugate in N ( P ) b y L e m m a 5.12. B u t x is central i n N ( P ) , and thus x is also central i n this group, and hence is conjugate o n l y to itself i n N ( P ) . It follows t h a t x = tx^t' for a l l t e T , and thus v { x ) = x \ \ because Y,nt = \ G : P \ . T h u s 1 = a ; l l , and x has p'-order. Since x P , we conclude t h a t x = 1, as desired.
0 t 1 n t 1 G n t G G G n t 1 G : P 0 G : P

A s a n a p p l i c a t i o n of T h e o r e m 5.18, we offer the following. 5.19. C o r o l l a r y . Suppose t h a t a S y l o w 2 - s u b g r o u p of a n o n a b e l i a n finite g r o u p G i s a d i r e c t p r o d u c t of c y c l i c s u b g r o u p s , one of w h i c h i s s t r i c t l y l a r g e r t h a n a l l of t h e o t h e r s . T h e n G i s n o t s i m p l e . P r o o f . L e t P e S y l ( G ) . W e can write P = A x P , where A is cyclic of order a > 2, and where W = 1 for a l l x e B . L e t C = { x / \ x G P } , and observe t h a t C is a n o n t r i v i a l characteristic subgroup of P of order 2, and thus the unique nonidentity element t of C is central i n N ( P ) . It follows by T h e o r e m 5.18 t h a t t 0 G', and thus G' < G. If G is simple, t h e n G' = 1, a n d so G is abelian, and this is a contradiction.
2 2 a 2 G

We mention that it is k n o w n from a part of the simple group classification t h a t if a Sylow 2-subgroup of a simple group is abelian, then i n fact, it must be elementary abelian. In other words, a l l of its cyclic direct factors have order 2. Problems 5C 5 C . 1 . Show t h a t B u r n s i d e ' s n o r m a l p-complement theorem is a consequence of T h e o r e m 5.18. 5 C . 2 . Suppose t h a t U C V C P C G, where P is an abelian Sylow 2subgroup of G and U and V are characteristic i n P such t h a t \ V : U\ = 2. If G is simple, show t h a t \ G \ = 2. 5 C . 3 . L e t G be simple and have a n abelian Sylow 2-subgroup P of order 2 . Deduce that P is elementary abelian.
5

Problems

5C

163

5 C . 4 . Suppose t h a t all Sylow subgroups of G are cyclic. Show t h a t for every divisor m of | G | , there exists a subgroup of order m, and show that every two subgroups of order m a r e conjugate i n G.

5 C . 5 . L e t P S y l ( G ) and suppose t h a t A a n d B are G-conjugate n o r m a l subgroups of P . Show that A and B are conjugate i n N ( P ) . If A is characteristic i n P , deduce that A = B .
p G

N o t e . If A is characteristic i n P e S y l ( G ) , it is c e r t a i n l y possible for A to be G-conjugate to some subgroup B of P w i t h B ^ A . B u t a m o n g a l l G conjugates 73 of A t h a t are contained i n P , the o n l y one t h a t can be n o r m a l in P is A itself.
p

5 C . 6 . L e t W C P C G . W e say t h a t VP is weakly closed i n P w i t h respect to G if the o n l y G-conjugate of W contained i n P is W itself. N o w suppose t h a t P e S y l ( G ) and that W C P .
p

(a) If is weakly closed i n P w i t h respect to G , show that W is weakly closed i n Q for each Sylow p-subgroup Q of G t h a t contains W . (b) Show t h a t W is weakly closed i n P w i t h respect to G i f and o n l y if W is n o r m a l i n N ( P ) a n d it is also n o r m a l i n a l l Sylow p subgroups of G that contain it.
G

(c) Suppose t h a t W is weakly closed i n P w i t h respect to G , and let W C P C G . Show t h a t every G-conjugate of W contained i n P is a c t u a l l y P - c o n j u g a t e to W. (d) Suppose t h a t W C Z ( P ) a n d t h a t V7 is weakly closed i n P w i t h respect to G . Show t h a t N ( W ) controls G-fusion i n P .
G

5 C . 7 . L e t \ G \ = 3 - 5 - l l . Show that G has a n o r m a l Sylow 3-subgroup.

5 C . 8 . L e t p > 2 be the smallest p r i m e divisor of \ G \ , a n d suppose that \ G \ is not divisible by p . Show that G has a n o r m a l p-complement.
3

5 C . 9 . Suppose t h a t G is nonabelian simple a n d has even order not divisible by 8. Show t h a t \ G \ is divisible by 3. N o t e . There do exist nonabelian simple groups w i t h order not divisible by 3, but a l l such groups have order divisible b y 64. I n fact, the o n l y nonabelian simple 3'-groups are the S u z u k i groups S z ( q ) , where q = 2 and e > 1 is odd. T h e group S z { q ) has order q - ( q - l ) - ( g + 1).
e 2 2

164

5.

Transfer

5 C . 1 0 . Suppose t h a t G is simple and has a n abelian Sylow 2-subgroup of order 8. Show t h a t \ G \ has order divisible by 7. N o t e . T h e simple group P S L ( 2 , 8) of order 504 = 2 - 3 - 7 has an elementary abelian Sylow 2-subgroup of order 8. T h e same is true for the smallest J a n k o sporadic simple group J w h i c h has order 175560 = 2 - 3 - 5 - 7 - l M 9 .
3 u 3 2

5 C . 1 1 . L e t H be a H a l l subgroup of G, and assume t h a t H C Z ( N ( J i ) ) . Show t h a t G has a n o r m a l p-complement for every p r i m e divisor p of \ H \ .

G

H i n t . W o r k by i n d u c t i o n on \ G \ . Show t h a t if P is a Sylow p-subgroup of H a n d N ( P ) < G, then P is central i n N ( P ) . If P < G and Q is a Sylow subgroup of H w i t h N ( Q ) < G, show t h a t P is central i n N ( Q ) .
G G G G

5 C . 1 2 . Suppose t h a t G has a cyclic Sylow p-subgroup, and let N < G have index divisible by p . P r o v e t h a t N has a n o r m a l p-complement. 5C.13.
G

(Navarro) Suppose t h a t P is a Sylow subgroup of G such t h a t P =

N ( P ) . Show t h a t N G ( P ' ) has a n o r m a l p-complement. H i n t . W i t h o u t loss of generality, P ' < G. Show t h a t G is p-solvable, and let X be a H a l l p-subgroup. Show t h a t G = P ' N ( X )
G

and a p p l y D e d e k i n d ' s

lemma.

5D We have seen t h a t transfer theory can sometimes be used to prove that a group is not simple, a n d as we saw i n the previous section, the transfer into a Sylow p-subgroup m a y be sufficient to accomplish this, especially w h e n the Sylow subgroup is abelian. W e ask now, w h e n does this work? I n other words, i f v is the transfer from G to a not necessarily abelian Sylow p-subgroup P of G, we want to know w h e n it is true t h a t k e r { v ) < G. Since G / k e r ( v ) = v { G ) C P / P ' , we see that G / k e r ( v ) is an abelian p-group, and hence transfer to a Sylow p-subgroup can prove n o n s i m p l i c i t y only for groups t h a t have factor groups t h a t are n o n t r i v i a l abelian p-groups. One of the results we prove i n this section is that if G a c t u a l l y has such a factor group, then this fact can definitely be established b y considering the transfer to a Sylow p-subgroup. In order to make a more precise statement, we introduce some notat i o n . G i v e n a p r i m e p and a finite group G, let AP(G) denote the unique smallest n o r m a l subgroup of G for w h i c h the corresponding factor group is an abelian p-group. ( T h a t A*(G) is well defined is an immediate consequence of the fact t h a t X M , N < G and b o t h G / M and G / N are abelian p-groups, then G / ( M n N ) is also an abelian p group.) T h e r e is, of course,

5D

165

an obvious analogy between the definitions of A*(G) a n d O { G ) , w h i c h , we recall, is the unique smallest n o r m a l subgroup of G whose factor group is a (not necessarily abelian) p-group. and G / A { G ) G'O {G).
p p

I n fact, A

{ G ) a n d O { G ) are more W e have O ( G ) C A*(G)

p

i n t i m a t e l y connected t h a n merely b y analogy.

AP(G)/OP(G) =

is isomorphic to the largest abelian factor of (G/CF(G))' = G ' O { G ) / O { G ) ,

P p p p

G/OP(G).

Thus =

a n d therefore, = O {G/G')

A*(G)

A l s o , it should be clear that A ( G ) / G ' G/G'.

is the n o r m a l

p-complement of

5 . 2 0 . T h e o r e m . L e t v : G -+ P / P ' be the t r a n s f e r h o m o m o r p h i s m , G is a finite g r o u p a n d P G S y l ( G ) . Then

p

where

ker(v) = A ( G ) .

In the previous section, we were able to o b t a i n results about the transfer from G into a n abelian Sylow subgroup P because i n t h a t case, N ( P )
G

controls G-fusion i n P . I n order to s t u d y the general case, where P is not necessarily abelian, we need other techniques t h a t help i n the study of fusion, a n d so we make the following definition, w h i c h is the key to the proofs of T h e o r e m 5.20 a n d other results i n this section. If H C G is an a r b i t r a r y subgroup, we define the f o c a l s u b g r o u p of H w i t h respect to G to be the subgroup generated by a l l elements of the form x~ y, If x , h G H , t h e n of course, x a n d x
X h x

where x a n d y are
G

G-conjugate elements of P . T h i s subgroup is denoted F o c ( P ) . are G-conjugate elements of H

h

(because t h e y are P - c o n j u g a t e ) , a n d so - x
G

lies i n F o c ( P ) . B u t x~ x
G

= Now

[ x , h ] , a n d this shows t h a t F o c ( P ) contains a l l c o m m u t a t o r s of elements of P . It follows t h a t H ' C F o c ( P ) , a n d i n fact, H ' = oc (H).

G H 9 x

assume x, y G P are G-conjugate, so t h a t y = x T h e n x~ y

G

for some element g G G .

= [ x , g ] G G ' , a n d hence F o c ( P ) C G'. F i n a l l y , we recall t h a t

if P C AT C G a n d AT controls G-fusion i n P , t h e n two elements x, y G P are G-conjugate if a n d only if they are AT -conjugate. It follows i n this case that F o c { H )
K

a n d F o c ( P ) have e x a c t l y the same generating sets, and so

G G

FOCK(P) = F o c ( P ) . T h r o u g h o u t m u c h of the rest of this chapter, our p r i m a r y interest w i l l be the transfer into a Sylow p-subgroup of G . M a n y of our arguments, however, w o u l d work as well for an a r b i t r a r y H a l l rr-subgroup, where TT is a set of primes. T h e following "focal subgroup theorem", for example, is v a l i d for an a r b i t r a r y H a l l yr-subgroup, a n d the p r o o f of the more general version goes t h r o u g h w i t h o u t change. define A
n

T h e o n l y e x t r a work required is to

{ G ) i n the obvious way: it is the unique smallest n o r m a l subgroup

of G for w h i c h the factor group is an abelian rr-group. 5.21. of a T h e o r e m (Focal S u b g r o u p ) . Suppose finite t h a t P i s a Sylow p-subgroup

g r o u p G, a n d l e t v : G - > P / P ' be the t r a n s f e r

homomorphism.

166

5.

Transfer

Then F o c ( P ) = P n G' = P n A { G ) = P n k e r ( u ) .
p G

P r o o f . F i r s t , observe that F o c ( P ) C P n G ' C P n A*(G)

G

C P n ker().
G

The first of these containments holds because the focal subgroup F o c ( P ) is contained i n P b y definition, and we observed previously t h a t F o c ( P ) C G'. T h e second containment holds since G / A { G ) is abelian, and so G' C A P ( G ) , and finally, the t h i r d containment holds because G / k e r ( v ) = v ( G ) C P / P ' , w h i c h is an abelian p-group, and thus A P ( G ) C ker(v). T o complete the proof, therefore, it suffices to show that P n ker(v) C F o c ( P ) .
G P G

Let x e P n k e r ( v ) , and let T be a right transversal for P i n G . T o apply the transfer-evaluation l e m m a to compute v ( x ) , we choose a n appropriate subset T C T a n d positive integers n for t G To. T h e n x and t x t ~ are G-conjugate members of P , a n d thus
n t n t l 0 t

{x-

){tx

r )

GFoc (P)
G

for a l l i e Tb. It follows t h a t the product of these elements (in any order) lies i n F o c ( P ) . A l s o , because P ' C F o c ( P ) , we can a r b i t r a r i l y rearrange elements of P i n this product to deduce that
G G

TT t%To or equivalents, Y[
ter
0

x~

n t

TT t J o

tx

n t

t-

e Foc (P),
G

tx n~

EE ] T x
ter
0

mod F o c ( P ) .
G

B y the transfer-evaluation l e m m a , the product o n the left is congruent m o d u l o P ' to V { x ) , where V is a pretransfer m a p from G to P . Since x G ker(u), we have V { x ) G P ' , and thus this product lies i n P ' C F o c ( P ) . It follows t h a t the product on the right also lies i n F o c ( P ) . B u t ^ n = \ G : P | , and this yields x \ \ G F o c ( P ) . Since | P | a n d \ G : P \ are coprime, there exists an integer m such t h a t m | G : P | EE 1 m o d | P | , and therefore,
G G t G : P G x

= (xl

G : P

l) GFoc (P),
G

as required.

T h e o r e m 5.20 is a n immediate consequence of the focal subgroup theorem. P r o o f of T h e o r e m 5 . 2 0 . W r i t e AT = ker(v) and A = A { G ) , so t h a t A C K , and our goal is to prove that equality holds here. Since \ G : K \ and
P

5D

167

| G : A \ are p-powers, it follows t h a t P K = G = P A , and by the focal subgroup theorem ( T h e o r e m 5.21), we have P n K = P n A . T h e n \G : K \ = \ P: P D K \ = \P:P H A \ = \G : A \ ,

a n d since A C AT, it follows t h a t A = AT, as wanted.

Now fix a Sylow p-subgroup P of G. L e t P C H C G , and let 1; and w, respectively, be the transfer h o m o m o r p h i s m s from G a n d P to P / P ' . W e w i s h to compare the kernels A ( G ) and A ( P ) a n d the images v { G ) and w(P) of these two transfer maps. Since P A ( G ) = G, we have P A ( G ) = G , and thus P / ( P n A ( G ) ) = G / A ( G ) is a n a b e l i a n p-group. It follows t h a t A P { H ) C P n A ( G ) , and thus | v ( G ) | = | G : A ( G ) | < | P : A ( P ) | = \ w { H ) \ . It should be clear that equality holds here i f and o n l y if A ( P ) = P n A P { G ) , and also if and o n l y if \ G : A P ( G ) \ = \ H : A ( P ) | . If equality does h o l d , we say that H controls p -transfer i n G . (Occasionally, we w i l l neglect to m e n t i o n the group G , and s i m p l y say t h a t " P controls ptransfer".)
P P P P P P P P P P P

T h e point here is this. If we know that P controls p-transfer i n G , a n d we also k n o w t h a t A ( P ) < H , i t follows t h a t A * ( G ) < G , and so G is nonsimple (unless | G | = p). In other words, a result t h a t guarantees t h a t some subgroup P containing a Sylow p-subgroup of G controls p-transfer allows us to shift the burden of p r o v i n g the existence of a n o n t r i v i a l abelian p-factor group from G to the smaller group P .
P

A s before, suppose that P C H C G , where P e S y l ( G ) . One way to guarantee t h a t P controls p-transfer i n G is to show t h a t P controls G fusion i n P . ( B u t we stress t h a t P can c o n t r o l transfer w i t h o u t c o n t r o l l i n g fusion.)
p

5.22. C o r o l l a r y . L e t P e S y l ( G ) , w h e r e G i s a finite g r o u p , a n d suppose that P C H C G , where H is a subgroup that controls G-fusion i nP . Then H controls p-transfer, and hence
p

A (P)

= P n A (G)
p

and

G/A (G) = P/A (P).

P r o o f . L e t v and w, respectively, be the transfer maps from G and from P to P / P ' . O u r goal is to show t h a t \ G : k e r ( v ) \ = | P : ker(to)|, or e q u i v a l e n t ^ (by two applications of L e m m a 5.11) t h a t | P : P n k e r ( w ) | = | P : Pnker(ti>)|. But P n k e r ( u ) = F o c ( P ) and P n k e r H = F o ( P ) by the focal subgroup theorem (5.21), and we have F o c ( P ) = F o c ( P ) since we are assuming that P controls G-fusion i n P . T h e result now follows. B
G C / / G H

5.23. C o r o l l a r y . L e t G be a finite g r o u p , a n d suppose a b e l i a n . Then N ( P ) controls p-transfer.

G

t h a t P e Syl (G) i s
p

168

5.

Transfer

P r o o f . W e k n o w by L e m m a 5.12 that N ( P ) controls G-fusion i n P . T h e result follows b y the previous corollary.

G

In fact, a c o n d i t i o n m u c h weaker t h a n t h a t the Sylow p-subgroup P is abelian is sufficient to establish that N ( P ) controls p-transfer. A theorem of P . H a l l a n d H . W i e l a n d t asserts that N ( P ) controls p-transfer whenever P has nilpotence class less t h a n p. (In other words, it suffices that P should not be too severely nonabelian.) A c t u a l l y , the H a l l - W i e l a n d t theorem is stronger t h a n we have just stated, a n d i n C h a p t e r 10, we present an even stronger result due to T . Y o s h i d a .
G G

It is not h a r d to deduce the B u r n s i d e n o r m a l p-complement theorem from C o r o l l a r y 5.23, and i n fact, the more general T h e o r e m 5.18 follows fairly easily too. A n o t h e r a p p l i c a t i o n of C o r o l l a r y 5.22 is the following. 5 . 2 4 . T h e o r e m . L e t G be a a m a x i m a l subgroup. VP r o o f . W e derive a contradiction by assuming t h a t p and q are distinct primes t h a t divide | P | . L e t P S y l ( P ) , and observe t h a t P < H since P is nilpotent. A l s o , P is n o n t r i v i a l , and P is not n o r m a l i n G since G is simple. T h e n P C N ( P ) < G , and we have H = N ( P ) by the m a x i m a l i t y of H . We deduce t h a t P is a full Sylow p-subgroup of G since otherwise, P < S for some p-subgroup S of G, a n d thus P < N ( P ) = Sn N ( P ) = S n P . T h i s is impossible, however, since S n H is a p-subgroup of P t h a t properly contains the Sylow p-subgroup P of P .
p G G 5 G

finite

s i m p l e g r o u p , a n d suppose

that H C G is prime

If H i s n i l p o t e n t , t h e n H i s a p - g r o u p f o r some

Similarly, P = N ( Q ) , where Q S y l ( P ) , a n d we have Q G S y l ( G ) . It follows by L e m m a 5.12 applied to Q that P = N ( Q ) controls G-fusion in C ( Q ) . B u t P C C ( Q ) since P and Q are n o r m a l subgroups of P and P n Q = 1. W e conclude t h a t H controls G-fusion i n P . Since P e S y l ( G ) and H = N ( P ) , we deduce from C o r o l l a r y 5.22 t h a t H controls p-transfer in G .
G ? ? G G G p G

Now P is nilpotent and has order divisible by p, and thus ( F ( P ) < P . It follows t h a t < P because the n o n t r i v i a l p-group must have a n o n t r i v i a l abelian h o m o m o r p h i c image. Since H controls p-transfer in G and A ( P ) < P , we have A ? ( G ) < G , and hence since G is simple, AP(G) = 1. T h e n G is a p-group, a n d this is the desired contradiction since q divides | G | .

AP(P)

P/OP(P)

The obvious question at this point is whether or not a nonabelian simple group a c t u a l l y can have a m a x i m a l subgroup t h a t is a p-group for some p r i m e p. T h e answer is "yes", but o n l y if p = 2. T h e Sylow 2-subgroup of

Problems

5D

169

the simple group PSL{2,17)

of order 2 - 3 - 1 7 is m a x i m a l , a n d the impossi-

b i l i t y for o d d primes is a consequence of T h o m p s o n ' s n o r m a l p-complement theorem, w h i c h we prove i n C h a p t e r 7. We can close this section w i t h a brief discussion of Tate's theorem, w h i c h be viewed as a k i n d of "supercharger" for transfer theory. Suppose t h a t ,
P

as usual, we have P C H C G , where P e S y l ( G ) . W e have seen t h a t under suitable hypotheses, it m a y be possible to show that H controls p-transfer, w h i c h , we recall, means t h a t A P ( H ) = P n A ? ( G ) . (We k n o w , for example, Tate's theorem t h a t this happens i f H controls G -fusion i n P , a n d there are also other situations where c o n t r o l of p-transfer c a n be established.) says t h a t i f A P { H ) A (H) case. G'
P P

= H n

A ( G ) , then, i n fact, ( F ( P ) = H f ) ( F ( G ) . ( T h e
p

converse of this is easy: if C F ( P ) = P n A (G).)

= P n C F ( G ) , it is routine to show t h a t

In order to see the significance of Tate's theorem, we consider a special Suppose that i n some group G, we have G ' f l P = P ' , where P G
P

S y l p ( G ) . Since | A ( G )
P

: G ' | is a p'-number, it follows easily t h a t A ( G ) n P =

P P

n P , a n d thus A ( G ) n P = P ' = A { P ) .

W e can now a p p l y Tate's = 1. It follows = P',

theorem w i t h H = P to conclude t h a t ( F ( G ) D P = O P ( P ) that

|OP(G)| is not divisible by p, a n d thus

P

C F ( G ) is a n o r m a l p-complement

in G . T o summarize, we have shown using Tate's theorem t h a t if G ' n P where P G S y l ( G ) , t h e n G must have a n o r m a l p-complement.

Unfortunately, we cannot prove Tate's theorem here. T h e r e seems to be no easy proof of this result, or even of the corollary t h a t we discussed i n the previous paragraph. Tate's original argument used cohomology theory, b u t we prefer T h o m p s o n ' s pretty character theoretic proof, w h i c h is presented in the author's character theory text.

Problems

5D

5 D . 1 . L e t P G S y l p ( G ) , where P is abelian. L e t P C H C G , a n d assume t h a t H controls p-transfer i n G . If H has a n o r m a l p-complement, show t h a t G has a n o r m a l p-complement. 5 D . 2 . L e t P G S y l ( G ) a n d assume t h a t P C Z ( G ) . W i t h o u t using B u r n P

side's n o r m a l p-complement theorem or any other part of transfer theory, deduce t h a t G has a n o r m a l p-complement. Use this to derive B u r n s i d e ' s theorem from C o r o l l a r y 5.23. 5 D . 3 . L e t G be a nonabelian simple group, a n d suppose t h a t P C G is a p-subgroup t h a t is a m a x i m a l subgroup of G . (a) Show t h a t P S y l ( G ) .
P

170

5.

Transfer

(b) If 1 < N < P a n d P / N is abelian, show t h a t P is the unique Sylow p-subgroup of G t h a t contains N . Deduce t h a t N is weakly closed in P w i t h respect to G. (See P r o b l e m 5C.6.) (c) Show t h a t the nilpotence class of P must be at least 3. H i n t . If P has class less t h a n 3, a p p l y (b) w i t h N = Z ( P ) , and use P r o b lem 5C.6(d).
p

5 D . 4 . L e t P G S y l ( G ) and suppose t h a t P C K C G. Show that O P { K ) C A" n O P ( G ) and t h a t if equality holds, then A ? ( K ) = AT n A ( G ) .

P

5 D . 5 . Suppose t h a t P e S y l ( G ) , and assume t h a t A l l P = P ' , where A = A P { G ) . If P ' < A , prove (without appealing to Tate's theorem) that G has a n o r m a l p-complement.
p

H i n t . U s i n g the Schur-Zassenhaus theorem, let i f b e a complement for P ' in A and show t h a t G = N ( K ) P ' . U s i n g the fact t h a t P ' C deduce t h a t P normalizes AT.
G

5 D . 6 . A s i n the previous problem, suppose that P G S y l ( G ) , and assume t h a t A n P = P ' , where A = A ( G ) . If P ' is abelian, prove (without appealing to Tate's theorem) that G has a n o r m a l p-complement.
p P

H i n t . L e t N = N ( P ' ) and observe that N satisfies the hypotheses. B u r n s i d e ' s n o r m a l p-complement theorem i n the group A .
G

Use

5E We return now to the p r o b l e m of determining when a finite group G has a n o r m a l p-complement. O f course, B u r n s i d e ' s theorem guarantees the existence of such a complement if P C Z ( N ( P ) ) , where P is a Sylow p subgroup of G. B u t since this hypothesis can o n l y h o l d if P is abelian, B u r n s i d e ' s result tells us n o t h i n g if P is nonabelian. T h e focal subgroup theorem, however, can be used to establish a necessary and sufficient cond i t i o n for G to have a n o r m a l p-complement, w i t h no assumption t h a t the Sylow subgroup is abelian.
G

5.25. T h e o r e m . A finite g r o u p G has a n o r m a l p - c o m p l e m e n t if a S y l o w p - s u b g r o u p of G c o n t r o l s i t s o w n f u s i o n i n G.

if a n d

only

P r o o f . F i r s t , assume t h a t N is a n o r m a l p-complement i n G, and let P G S y l ( G ) . T h e n N P = G and N n P = 1, and thus the restriction to P of

p

the canonical h o m o m o r p h i s m G - > G = G / N G/N. N o w suppose t h a t x,y

is an i s o m o r p h i s m of P onto

G P are conjugate i n G. T h e n their images x

and y are conjugate i n G, a n d it follows v i a the i s o m o r p h i s m u h > u from P

5E

171

to G t h a t x and y are conjugate i n P . T h u s P controls its own fusion i n G, as required. Conversely, assume t h a t P controls its o w n fusion i n G. L e t N = O ( G ) , a n d w r i t e Q = N n P , so that Q G S y l ( / V ) . W e w i l l show t h a t TV is a n o r m a l p-complement i n G by p r o v i n g t h a t Q = 1. W e argue first t h a t A ( N ) = N . To see w h y this is so, observe t h a t A ( N ) is characteristic i n N , a n d hence is n o r m a l i n G . Furthermore,
p P P p

\G :A { N ) \

= \ G : iV||/V : A ( N ) \ ,
p

w h i c h is a power of p , and thus N = O ( G ) C A * (AT). It follows that i V = A { N ) , as c l a i m e d . B y the focal subgroup theorem, therefore, Focyv(G) = Q n A { N ) = Q n N = Q.
p p

N o w let x,y G Q be TV-conjugate, so t h a t x " ? / is a t y p i c a l generator for FOCJV(Q). T h e n x and y are G-conjugate elements of P , a n d so by hypothesis, they are conjugate i n P , and we c a n w r i t e y = x w i t h u P . W e have x~ y = x ~ x = [ x , u ] G [ Q , P ] , a n d it follows t h a t F o c i v ( g ) C [ Q , P ] . Since F o c { Q ) = Q, we have Q C [ Q , P ] , and thus Q C [Q, P , P , . . . , P ] , where there are a r b i t r a r i l y m a n y commutations w i t h P . B u t P is nilpotent a n d Q C P , a n d thus w i t h sufficiently m a n y c o m m u tations, we have [Q, P , P , . . . , P ] = 1. It follows t h a t Q = 1, and so i V is a n o r m a l p-complement i n G , as wanted.
u l 1 u N

A l t h o u g h it is pleasant that T h e o r e m 5.25 gives a necessary and sufficient c o n d i t i o n for G to have a n o r m a l p-complement, this result is of l i m i t e d a p p l i c a b i l i t y because i n general, it is difficult to determine whether or not a Sylow p-subgroup of G controls its o w n fusion. A deeper and m u c h more useful c r i t e r i o n for G to have a n o r m a l p-complement is the following result of Frobenius, w h i c h shows that the existence of a n o r m a l p-complement i n G is determined by the p-local subgroups of G . (Recall that a subgroup N C G is said to be p-local i n G if N = N ( A ) , where X is some nonidentity p subgroup of G . )
G

5 . 2 6 . T h e o r e m (Frobenius). L e t G be a p r i m e . T h e n the following are equivalent. (1) G has a n o r m a l p - c o m p l e m e n t . (2) N

G

finite

g r o u p , a n d suppose

p is a

p O has a n o r m a l p - c o m p l e m e n t f o r every i s a p - g r o u p f o r every

nonidentity

p-subgroup

I C G . (3) N ( X ) / C ( X )
G G

p-subgroup I C G .

Observe that i f we were to drop the w o r d "nonidentity" from assertion (2), the theorem w o u l d still be true, but it w o u l d be far less interesting. T h i s is so because if X = 1, then N ( A ) = G , a n d the i m p l i c a t i o n (2) (1)
G

172

5.

Transfer

w o u l d have no content. O n the other hand, assertion (3) is a u t o m a t i c a l l y true when X = 1, a n d hence it is irrelevant whether or not we restrict (3) to nonidentity subgroups X . (We have chosen not to include such a restriction.) A l s o , we stress t h a t i n general, it is not true t h a t if X C G is a p-subgroup and N ( X ) / C ( X ) is a p-group, then N ( A ) has a n o r m a l p-complement. T h e theorem tells us t h a t if for every p-subgroup X Q G , \ t is true t h a t N { X ) / C ( X ) is a p-group, then it is also true that N ( A ) has a n o r m a l p-complement for every p-subgroup I C G . T h e i m p l i c a t i o n (3) => (2) is not v a l i d for i n d i v i d u a l p-subgroups X , however.
G G G G G G

T h e i m p l i c a t i o n s (1) (2) => (3) of Frobenius' theorem are nearly t r i v i a l , a n d we dispose of t h e m first. 5 . 2 7 . L e m m a . I n t h e s i t u a t i o n of T h e o r e m 5.26, if { I ) h o l d s , t h e n so (2), a n d if ( 2 ) h o l d s , t h e n so does (3). does

P r o o f . T o prove t h a t (1) => (2), we show more: i f G has a n o r m a l pcomplement, then every subgroup H C G has a n o r m a l p-complement. T o see this, suppose t h a t N is a n o r m a l p-complement i n G. T h e n J V f l / i i s a n o r m a l subgroup of H h a v i n g p'-order, and since \ H : H n N \ = \ N H : N \ divides \ G : N \ , w h i c h is a p-power, it follows that N n H is actually a n o r m a l p-complement i n H , as wanted. N o w assume (2), and let X C G be a p-subgroup. If X = 1, then N ( A ) / C ( A ) is t r i v i a l , and so is a p-group, as required. Otherwise, we know that N ( A ) has a n o r m a l p-complement AT, a n d we see that K centralizes X since X n AT = 1 and b o t h X and K are n o r m a l subgroups of N ( A ) . T h e n AT C C ( A ) , and so | N ( A ) : C ( A ) | divides | N ( A ) : AT|, w h i c h is a power of p. T h u s N ( A ) / C ( A ) is a p-group, as required.
G G G G G G G G G G

T h e key to the proof of F r o b e n i u s ' theorem is the following. 5 . 2 8 . L e m m a . Assume t h a t N { X ) / C { X ) i s a p - g r o u p f o r every p - s u b g r o u p X of a f i n i t e g r o u p G, a n d l e t P , Q S y l ( G ) . T h e n Q = P , f o r some element c e C (PnQ).
G G c p G

5E

173

P r o o f . L e t V be the set of pairs ( P , Q ) , necessarily distinct, a n d such that Q = P

c

where P , Q S y l ( G ) are not

p

for some element c G C ( P n Q ) .

G

O u r goal, of course, is to show that the set V consists of a l l pairs of members of S y l p ( G ) . If this is false, choose P , Q G S y l ( G ) w i t h P n Q as large as
p

possible such t h a t ( P , Q ) 0 V. W r i t e D = P n Q, a n d observe t h a t D < P a n d D < Q , since otherwise P = D = Q, a n d thus ( P , Q ) lies i n P since we can take c = 1. Let /V = N ( F > ) . Since F n N a n d Q n N are p-subgroups of TV, they
G

are contained i n Sylow p-subgroups S a n d T of N , respectively, a n d S is contained i n some Sylow p-subgroup R , of G as shown i n the d i a g r a m . L e t C = C ( D ) , so t h a t i V / C is a p-group by hypothesis. T h e n N = SC since
G

S is a Sylow p-subgroup of N . B y the Sylow C - t h e o r e m , T = S T = S D We

n

for some Then

element n N , a n d we can w r i t e n = sy, where s G 5 a n d y C. = S

sy

= S,

a n d since 5 C H , it follows t h a t T = S

C Ry.
P

N o w P (~) N C S C R, a n d s o P f l J V C P n

But P H N = N ( > ) >

since D < P a n d "normalizers grow" i n p-groups. can thus w r i t e R = P

x

W e conclude t h a t ( F n F ) , and i n
y

P (1 R > D = P H Q , a n d so by the choice of F a n d Q, we have ( F , F ) e F . for some element a; e C

G

p a r t i c u l a r , we observe that x centralizes D . A l s o , Q n TV C T C R , a n d thus Q n i V C F n g . R

y

B u t Q n J V = N Q ( D ) > F> since D < Q, a n d thus

y z

n Q

> D = P n Q . A g a i n by the choice of F a n d Q, we have ( F ^ , Q ) F ,

G x y z

a n d we can write Q = ( R ) for some element z G C ( F ^ n Q ) , a n d so i n p a r t i c u l a r , z centralizes D . N o w P i n F , w h i c h is a c o n t r a d i c t i o n . P r o o f of T h e o r e m 5.26. = R

y z

= Q, a n d since each of x, y

a n d z centralizes D , it follows that x y ^ centralizes D , a n d thus ( F , Q ) lies

B y L e m m a 5.27, it suffices to prove t h a t (3)

(1), a n d so we assume (3), a n d we show t h a t G has a n o r m a l p-complement. B y T h e o r e m 5.25, it is enough to prove t h a t a S y l o w p-subgroup F of G controls its o w n fusion i n G . W e therefore assume t h a t x,y W r i t e y = x w i t h g G G , a n d observe t h a t y G P n P .
9

G P are conjugate B y L e m m a 5.28,

in G , a n d we work to show that x a n d y are a c t u a l l y conjugate i n F . there is an element c G C ( F n P ) such t h a t ( P ) = P , a n d we see t h a t

9 9 c G G

c centralizes y . B y hypothesis, N ( F ) / C ( F ) is a p-group, a n d since F

174

5.

Transfer

is a Sylow p-subgroup of N ( P ) , we can write N ( P ) = C ( P ) P . W e have #c G N ( P ) , and so we can w r i t e gc = t u , where t G C ( P ) and u G P . In p a r t i c u l a r , since x G P , we see that t centralizes x , and thus = y = = x = x . Since u G P , it follows that x and y are conjugate in P , as required. T h i s completes the proof.
G G G G G c g c t u u y x

H o w can we check that N ( X ) / C { X ) is a p - g r o u p for some p-group X I T h i s is equivalent to saying that C ( X ) contains a full Sylow g-subgroup of N ( X ) for every p r i m e q different from p. In other words, a group G satisfies (3) of F r o b e n i u s ' theorem if whenever a g-subgroup Q of G acts on (normalizes) a p-subgroup X of G, the a c t i o n is t r i v i a l . T h i s yields the following corollary.
G G G G

5.29. C o r o l l a r y . Suppose G i s a finite g r o u p a n d \ G \ = p m , w h e r e p does n o t d i v i d e m. Assume t h a t n o p r i m e d i v i s o r q o f m d i v i d e s a n i n t e g e r of t h e f o r m p - 1 , w h e r e 1 < e < a . T h e n G has a n o r m a l p - c o m p l e m e n t .

e

It follows, for example, that every nonabelian simple group of even order either has order divisible by 32, or by one of 3, 5 or 7. T o see why this is so, write \ G \ = 2 m , where TO is o d d , and suppose t h a t a < 5. Since G is nonabelian and simple and has even order, it does not have a n o r m a l 2-complement. B y C o r o l l a r y 5.29, therefore, m cannot be coprime to a l l of the numbers 2 - 1, where 1 < e < 4. Since 2 - 1 = 1, 2 - 1 = 3, 2 - 1 = 7 and 2 - 1 = 15, at least one of the primes 3, 5 or 7 must divide m.
a e 1 2 3 4

O f course, by the F e i t - T h o m p s o n odd-order theorem, every nonabelian simple group has even order. A l s o , it follows from the classification of simple groups t h a t the order of every such group is divisible by 3 or 5. P r o o f of C o r o l l a r y 5.29. If G does not have a n o r m a l p-complement, t h e n (3) of F r o b e n i u s ' theorem must fail, and thus G contains a g-subgroup Q t h a t acts n o n t r i v i a l l y o n some p-subgroup X , where q is some prime d i visor of TO. N o w C ( Q ) < X , so we can write \ X : C ( Q ) \ = p , where 1 < e < a . A l l of the elements of X - C ( Q ) lie i n Q-orbits of size divisible by q, and it follows that q divides \X\ - \ C { Q ) \ = \ C ( Q ) \ { p - 1). B u t | C x ( Q ) | is a power of p and hence is coprime to q, and thus q divides p - 1. The result now follows.
e X X X e X x e

A somewhat less t r i v i a l a p p l i c a t i o n is the following. 5.30. C o r o l l a r y . L e t p be a n o d d p r i m e , a n d suppose t h a t every element of o r d e r p i n t h e finite g r o u p G i s c e n t r a l . T h e n G has a n o r m a l p - c o m p l e m e n t . P r o o f . Otherwise, G has a p-subgroup X acted on n o n t r i v i a l l y by a q-

subgroup Q, where q is some p r i m e different from p. Since the elements of

Problems 5 E

175

order p i n A are central i n G, a l l of t h e m are fixed by Q, and it follows by T h e o r e m 4.36 t h a t Q acts t r i v i a l l y o n X . T h i s is a contradiction. A l t h o u g h C o r o l l a r y 5.30 fails for p = 2, it is true that G has a n o r m a l 2-complement if every element x e G such t h a t x = 1 is central i n G. T h e proof of this is essentially the same as t h a t of C o r o l l a r y 5.30, except that one must appeal to P r o b l e m 4 D . 4 i n place of T h e o r e m 4.36.
A

W e m e n t i o n that if p > 2, there is a m u c h stronger version of the i m p l i c a t i o n (2) (1) i n Frobenius' theorem. I n order to show that G has a n o r m a l p-complement, it is not necessary to check t h a t N ( A ) has a n o r m a l p-complement for every nonidentity p-subgroup X of G. In fact, it suffices to consider just t w o such subgroups: the center Z ( P ) , and a certain characteristic subgroup J ( P ) , where P e S y l ( G ) . T h i s theorem of T h o m p s o n is proved i n C h a p t e r 7.
G p

Problems 5E 5 E . 1 . L e t G be finite, and suppose that every proper subgroup of G has a n o r m a l p-complement but that G itself does not. Show t h a t G has a n o r m a l Sylow p-subgroup a n d that \ G \ has exactly one p r i m e divisor other t h a n p . 5 E . 2 . L e t G be a finite group i n w h i c h every proper subgroup is supersolvable. Show that G is solvable. H i n t . If G is a counterexample of m i n i m a l order, show that G is simple. T h e n use P r o b l e m 3 B . 1 0 to show t h a t G must have a n o r m a l p-complement, where p is the smallest prime divisor of \ G \ . 5 E . 3 . Suppose that every two-generator subgroup of a finite group G has a n o r m a l p-complement. Show t h a t G has a n o r m a l p-complement. H i n t . Consider ( x , y ) , where x lies i n some p-subgroup P and y e N ( P ) has order not divisible by p .
G

Chapter 6

Frobenius Actions

6A Let A and N be finite groups, and suppose t h a t A acts o n N v i a automorphisms. T h e action of A on N is said to be F r o b e n i u s i f n ^ n whenever n G N and a G A are nonidentity elements. E q u i v a l e n t l y , the action of A on N is Frobenius if and o n l y if C ( a ) = 1 for a l l n o n i d e n t i t y elements a G A , a n d also if and o n l y if C ( n ) = 1 for a l l n o n i d e n t i t y elements n G N . Y e t another point of view is to consider the orbits of the a c t i o n of A on the elements of N . If 1 + n G N , then by the fundamental c o u n t i n g p r i n c i p l e , the size of the A - o r b i t of n is \ A : C { n ) \ , and so we see that the a c t i o n is Frobenius if and only i f a l l A - o r b i t s on N other t h a n the t r i v i a l orbit {1} have size equal to | A | . (Recall that a n orbit of an action of a finite group A on some set is said to be regular if the orbit size is equal to | A | . If a l l A - o r b i t s are regular, the action is semiregular.) T h u s an action v i a automorphisms of A o n N is Frobenius precisely w h e n A acts semiregularly on the n o n i d e n t i t y elements of N . F i n a l l y , to conclude this i n t r o d u c t i o n to Frobenius actions, we observe that if A acts on N v i a automorphisms and the action is Frobenius, then the n a t u r a l a c t i o n of every subgroup of A on N is also Frobenius, a n d so is the action of A on every subgroup of N t h a t admits A .
a N A A

B y the following easy observation, Frobenius actions are a u t o m a t i c a l l y coprime actions. E v e r y t h i n g we k n o w about coprime actions from C h a p t e r 3, therefore, applies to Frobenius actions. 6.1. L e m m a . L e t A a n d N be finite g r o u p s , a n d suppose there is a Frobenius a c t i o n of A o n N . Then \ N \ = 1 m o d \ A \ , a n d hence \ N \ and \A\ a r e coprime. 177

178

6. F r o b e n i u s A c t i o n s

P r o o f . T h e set N is decomposed into A - o r b i t s . O n e of these is {1} and a l l of the others are regular orbits, w h i c h have size T h e result follows. 6 . 2 . C o r o l l a r y . L e t A a n d N be finite g r o u p s , a n d suppose there is a Froben i u s a c t i o n of A o n N . L e t M be a n A - i n v a r i a n t n o r m a l s u b g r o u p of N . T h e n t h e i n d u c e d a c t i o n of A o n N / M i s F r o b e n i u s . P r o o f . L e t 1 7^ a G A . T h e n ( a ) acts coprimely on N , and we recall from C o r o l l a r y 3.28 and the discussion preceding it that i n coprime actions, "fixed points come from fixed points" i n the induced action on the factor group N / M = N . (This assumes t h a t one of (a) or M is solvable, but since (a) is
F N

cyclic, this hypothesis is satisfied.) It follows t h a t C ( ( a = C { ( a ) ) , and this is t r i v i a l since C ( ( a ) )

N

= 1. T h u s C ( a ) = 1, a n d so the action of A
F

on N is Frobenius, as required.

Before we proceed to develop the theory of Frobenius actions, it seems appropriate to discuss some examples. O f course, if either A or N is the t r i v i a l group, then the unique (but not very interesting) action of A on N is Frobenius, but it is almost as easy to construct some n o n t r i v i a l Frobenius actions. F o r example, if A has order 2 w i t h nonidentity element a , and N is abelian of o d d order, we can define a Frobenius action of A on N by setting n = n for a l l n G N .
a _ 1

N e x t , consider a finite field F of order q. L e t N be the additive group of F , and let A be a subgroup of the m u l t i p l i c a t i v e group F = F - {0} of F . T h u s N is an elementary abelian p-group, where p is the unique p r i m e divisor of q, and since F is cyclic of order q - 1, we can take A to be cyclic of order r for a n a r b i t r a r y divisor r of q - 1. If n G N and a G A , we define an action by setting n = n a , where the product n a is denned by viewing b o t h n and a as elements of F and using the field m u l t i p l i c a t i o n . T h i s really is an action v i a automorphisms since if x , y G N , then ( x + y ) = { x + y ) a = x a + y a = x + y by the d i s t r i b u t i v e law. T o see that this action is Frobenius, suppose that n = n . T h e n n a = n , and so n ( a - 1) = 0. It follows t h a t either n = 0, w h i c h is the identity i n N , or a = 1, w h i c h is the identity i n A .
x x a a a a a

G i v e n primes p a n d r, let N and A , respectively, have orders p and r . B y the construction of the previous paragraph, we see that there is a Frobenius action of A o n N provided that r divides p - 1. ( O f course, this d i v i s i b i l i t y c o n d i t i o n is necessary by L e m m a 6.1.) I n fact, an a r b i t r a r y n o n t r i v i a l action of a group A of p r i m e order r on a group N of p r i m e order p is necessarily Frobenius since if n G N and a G A are nonidentity elements, then a cannot centralize n or else A = ( a ) centralizes N = ( n ) , and the action is t r i v i a l . In p a r t i c u l a r , if the action of A on N is n o n t r i v i a l , then r must divide p - I by L e m m a 6.1.

6A

179

Not every group A can have a Frobenius a c t i o n on a n o n t r i v i a l group N , and not every group N can admit a Frobenius a c t i o n of a n o n t r i v i a l group A . A n easy result of this type is the following, but as we shall see after we develop the necessary machinery, m u c h more is true. 6.3. T h e o r e m . L e t A a n d N be finite g r o u p s , a n d suppose there is a Froben i u s a c t i o n of A o n N . Assume t h a t \ A \ i s even a n d t h a t N i s n o n t r i v i a l . Then A contains a unique involution, and N is abelian. P r o o f . Since | A | is even by assumption, we can fix some i n v o l u t i o n t G A , a n d we consider the m a p x i-> x " V from N to itself. If x , y G N and ~ t = y - i t then y x ~ = ^ ( x ) " = ( y x " ) ' - Since the action of A is Frobenius and t ^ 1, it follows t h a t y x ' = 1, a n d thus x = y . T h i s shows t h a t our m a p x H- x ~ x is injective, a n d hence it is also surjective because N is finite. G i v e n an a r b i t r a r y element y G N , therefore, we can write y = x " V for some element xN. Thus
l l 4 1 1 x x y t 1 l l

= (x-V)' = (x-'Yx = (x'y'x = (x-V)"

2

= y -

where the second equality holds since t = 1. If s G A is an a r b i t r a r y i n v o l u t i o n , a similar c a l c u l a t i o n yields y = y ~ \ a n d thus y = y . T h e n ~ = y for a l l y N , and since N is n o n t r i v i a l by assumption, we can assume t h a t y ^ 1, a n d thus st' = 1 because the action of A is Frobenius. T h u s s = t, and hence t is the unique i n v o l u t i o n i n A , as required.
s s l s t x y 1

We have seen that t inverts every element of TV, and thus if x , y G N , then ( x y Y = ( x y ) ' = y ^ x ' = ? r V = ( y x ) K It follows t h a t x y = y x , and thus N is abelian.
1 1

A finite group A is said to be a F r o b e n i u s c o m p l e m e n t if it has a Frobenius action on some nonidentity group N , a n d similarly, a finite group N is a F r o b e n i u s kernel if it admits a Frobenius action by some nonidentity finite group A . ( O u r definition of "Frobenius complement" is not quite standard, but as we e x p l a i n later, it is equivalent to the usual definition.) B y T h e o r e m 6.3, we k n o w t h a t a Frobenius complement of even order contains a unique i n v o l u t i o n , a n d we w i l l prove t h a t a Frobenius complement of o d d order is even more t i g h t l y constrained. A n o d d order Frobenius complement, for example, contains a unique subgroup of order p for every p r i m e p d i v i d i n g its order. There are also strong s t r u c t u r a l constraints on Frobenius kernels, as is suggested by the fact t h a t the Frobenius kernel i n T h e o r e m 6.3 is abelian. In general, Frobenius kernels need not be abelian, but they are always nilpo tent. T h e nilpotence of Frobenius kernels h a d been a long open problem, w h i c h was finally settled by J . T h o m p s o n i n his P h . D . thesis i n 1959. W e start work toward a proof of T h o m p s o n ' s theorem i n this chapter, and we

180

6. F r o b e n i u s

Actions

complete the proof i n C h a p t e r 7, where we present one of the key ingredients: T h o m p s o n ' s deep n o r m a l p-complement theorem. A l l of the Frobenius complements a n d Frobenius kernels t h a t we have seen so far are abelian, and i n fact, a l l of our Frobenius complements have been cyclic. (We shall see that a n abelian Frobenius complement is necessarily cyclic.) B u t nonabelian examples of Frobenius complements a n d kernels exist too. F i r s t , to show t h a t a Frobenius kernel can be nonabelian, let F be a finite field of order q, and let N be the group of "unitary" upper triangular 3 x 3 matrices over F . (The elements of TV are a l l of the 3 x 3 matrices over F w i t h ones on the diagonal and zeros below the diagonal.) It should be clear that \ N \ = q , and it is easy to check that N is not abelian. If q - 1 is not a power of 2, we show that TV is a Frobenius kernel by constructing a nonidentity subgroup A of the group G L ( 3 , F ) of all invertible 3 x 3 matrices over F such t h a t A acts by conjugation on N , and we w i l l show that this action is Frobenius. (Note t h a t although N is not abelian, it is nilpotent, as is required by T h o m p s o n ' s theorem.)
3

Let C be a (necessarily cyclic) odd-order subgroup of the m u l t i p l i c a t i v e group F of F , and let A be the group of 3 x 3 diagonal matrices of the form a = d i a g ( l , c, c ) , where c runs over C . (Note that we can choose C so t h a t \ C \ is an a r b i t r a r y o d d divisor of | F | = q - 1, a n d so, i n particular, i f q - 1 is not a power of 2, we can suppose that A is n o n t r i v i a l . )
x 2 X

It is easy to check that A acts by conjugation on N , and t h a t i n fact, 1 0 0

2

"1

1 0

y i_

0 0

cx 1 0

c z~ cy 1

where a = d i a g ( l , c, c ) . If a is a nonidentity element of A , then c + 1, and thus also c ^ 1 because we are assuming that C has o d d order. If n = n i n this situation, we see that each of the above-diagonal entries x , y and z of n must be zero, and thus n = 1. It follows t h a t the action is Frobenius.
2 a

N e x t , we mention some examples of nonabelian Frobenius complements. F i r s t , consider the group S = 5 L ( 2 , 3 ) of 2 x 2 matrices w i t h determinant 1 over the field of order 3. It is easy to see that \S\ = 24, and it is not h a r d to check that S has a unique Sylow 2-subgroup, w h i c h we w i l l call Q. O f course, \ Q \ = 8, and i n fact, Q is isomorphic to the quaternion group Q . I n p a r t i c u l a r , Q is noncyclic, a n d its unique element of order 2 is the negative of the identity m a t r i x . Since Q is a subgroup of 5 L ( 2 , 3 ) , it has a n a t u r a l action o n the vector space V of dimension 2 over the field of order 3, and since the unique element of order 2 i n Q fixes no nonzero vector, it follows that the action of Q on the additive group of V is Frobenius. (Note that
8

6A

181

since \V\ = 1 + \ Q \ , the a c t i o n of Q o n the n o n i d e n t i t y elements of V is regular, a n d not just semiregular.) N e x t , we appeal to the not completely obvious fact t h a t S L ( 2 , 5) contains an isomorphic copy S of 5 L ( 2 , 3 ) . (In fact, S is the normalizer i n 5 L ( 2 , 5 ) of a Sylow 2-subgroup.) Since S C 5 L ( 2 , 5 ) , there is a n a c t i o n of S o n a 2-dimensional vector space W over a field of order 5, a n d it easy to see t h a t no element of order 2 or order 3 i n S L { 2 , 5) c a n have an eigenvalue equal to 1. N o such element, therefore, c a n have a n o n t r i v i a l fixed point on W, and it follows t h a t the action of S o n W is F r o b e n i u s . (Note t h a t i n this case too, the action on nonidentity elements is regular a n d not just semiregular since \W\ = 25 = 1 + 24 = 1 + \S\.) Somewhat similarly, the group 5 L ( 2 , 5 ) c a n be embedded i n SL(2,11), and this yields a Frobenius action of 5 L ( 2 , 5 ) o n a vector space of order l l = 121. (Since | 5 L ( 2 , 5 ) | = 120, we again have a regular action o n nonidentity elements.) B u t SL(2,11) is not a Frobenius complement, and so the story ends here. (In fact, SL(2,11) contains a nonabelian subgroup of order 55, but we w i l l prove t h a t no Frobenius complement can have a n o n c y c l i c subgroup whose order is a product of two primes.)
2

N o t i c e t h a t the Frobenius complement 5 L ( 2 , 5 ) is not solvable. I n fact, SL(2,5) is perfect, w h i c h , we recall, means t h a t it is its o w n derived subgroup. A theorem of H . Zassenhaus asserts t h a t S L { 2 , 5) is the o n l y perfect Frobenius complement. (We m e n t i o n t h a t one of the few serious errors i n W . B u r n s i d e ' s classic group theory book is his assertion t h a t Frobenius c o m plements are always nilpotent. T h i s false statement is his T h e o r e m V o n Page 336 of the 1911 edition.) N e x t , we consider the semidirect p r o d u c t G = N x A , where A acts on N , and we ask what can be said about G i f the given a c t i o n is Frobenius. A s usual when we consider semidirect products, we view N a n d A as subgroups of G, where N is n o r m a l a n d A is a complement for N i n G. (Recall t h a t this means t h a t N A = G and N n A = 1.) I n this s i t u a t i o n , of course, the original a c t i o n of A on N is exactly the conjugation a c t i o n of A on N i n G.

6.4. T h e o r e m . L e t N be a n o r m a l s u b g r o u p of a

finite

g r o u p G, a n d

suppose

that A is a complement f o r N i n G . The following are then (1) T h e c o n j u g a t i o n a c t i o n of A o n N i s F r o b e n i u s . (2) A C \ A

G 9

equivalent.

= l f o r a l l elements

g G - A . a e A . n e N .

(3) C ( a ) C A f o r a l l n o n i d e n t i t y elements (4) C ( n ) C N f o r a l l n o n i d e n t i t y elements

G

182

6. F r o b e n i u s

Actions

If b o t h N and A are n o n t r i v i a l i n the s i t u a t i o n of T h e o r e m 6.4, we say t h a t G is a F r o b e n i u s g r o u p and t h a t A and N are respectively the Frobenius complement a n d Frobenius kernel of G. We need the following easy c o m p u t a t i o n .

6.5. L e m m a . L e t A be a s u b g r o u p of a finite g r o u p G, a n d suppose that A n A = 1 f o r a l l elements g G G - A . L e t X be t h e subset of G c o n s i s t i n g of those elements t h a t a r e n o t c o n j u g a t e i n G t o any n o n i d e n t i t y element of A . T h e n \X\ = \ G \ / \ A \ .
9

P r o o f . If A = 1, then A has no nonidentity elements, and so X = G and there is n o t h i n g further to prove. A s s u m i n g t h a t A > 1, we see that A = N ( A ) since if A = A , then A n A = A > 1, and thus x G A . It follows t h a t A has exactly \ G : A \ distinct conjugates i n G. Furthermore, since each of these conjugates satisfies the c o n d i t i o n we assumed about A , no two of t h e m can have a n o n t r i v i a l intersection. T h e conjugates of A , therefore, account for a t o t a l of \ G : A \ { \ A \ - 1) nonidentity elements of G, a n d these are exactly the elements of G that are conjugate to nonidentity elements of A . W e conclude t h a t \X\ = \ G \ - \ G : A \ ( \ A \ - 1) = \ G : A \ , as required.
x x G

6.6. C o r o l l a r y . L e t N be a n o r m a l s u b g r o u p of a finite g r o u p G, a n d suppose t h a t A i s a c o m p l e m e n t f o r N i n G such that A n A = 1 f o rall elements g e G - A . T h e n N i s e x a c t l y t h e set X of L e m m a 6.5. I t i s , i n o t h e r w o r d s , t h e set of elements of G t h a t a r e n o t c o n j u g a t e t o any n o n i d e n t i t y element of A .
9

P r o o f . Since N n A = 1, it is also true that N n A = 1 for a l l g G G, and thus no element of N can be conjugate to a nonidentity element of A . It follows t h a t J V C ! B y L e m m a 6.5, however, \X\ = \ G \ / \ A \ = \ N \ , and the result follows.
9

In fact, the set X of L e m m a 6.5 is a l w a y s a subgroup. T h i s remarkable theorem of Frobenius was proved using character theory, and no one has ever found a character-free proof. (We mention t h a t it was Frobenius who invented character theory, and his proof that the set AT is a subgroup is one of its earliest applications.) Observe t h a t once we know t h a t A is a subgroup, it is immediate t h a t it is a n o r m a l subgroup. A l s o , b y the definition of X , we know t h a t X n A = 1, and so given that A is a subgroup, it follows by L e m m a 6.5 t h a t \ X A \ = \ X \ \ A \ = \ G \ , and thus A complements the n o r m a l subgroup X i n G. I n other words, we are i n the s i t u a t i o n of T h e o r e m 6.4, and c o n d i t i o n (2) of t h a t theorem holds. It follows t h a t (1) holds, and so the conjugation a c t i o n of A o n X is Frobenius. I n p a r t i c u l a r , if A < G, then A is a Frobenius complement i n G. ( O f course, the assumption t h a t A is

6A

183

proper is needed to guarantee that the group X is nontrivial.) A s s u m i n g Frobenius' theorem, therefore, we have shown t h a t a subgroup A < G is a Frobenius complement i n G if and o n l y if A D A = 1 for a l l elements g G G - A . I n fact, m a n y authors use this is as the definition of a Frobenius complement. ( A n d indeed, we presented this definition i n the note following p r o b l e m 1D.3.)
9

A n o t h e r way to t h i n k about F r o b e n i u s ' theorem is this. Suppose t h a t G acts t r a n s i t i v e l y on some set ft, a n d assume t h a t no nonidentity element of G fixes as m a n y as two members of ft. L e t a G ft, and let A = G , the stabilizer of a . N o w suppose t h a t g e G A , and write 3 = a-g, so t h a t p + a . T h e n A = G p , and so every element of A n A fixes b o t h a a n d p . B y assumption, therefore, A n A = 1, and thus the point stabilizer A = G satisfies the hypothesis of L e m m a 6.5. T h e elements of G conjugate to nonidentity elements of A are exactly the nonidentity elements of G t h a t fix some point of ft, and so i n this context, the set X of L e m m a 6.5 consists exactly of the identity together w i t h those elements of G t h a t fix no point of ft. F r o b e n i u s ' theorem tells us, therefore, t h a t i f G acts transitively o n a set ft and no nonidentity element of G fixes as m a n y as two points, then the set consisting of the identity and the elements of G t h a t fix no points forms a subgroup. I n fact, it is not h a r d to see t h a t this assertion about group actions is equivalent to Frobenius' theorem.
a 9 9 9 a

P r o o f o f T h e o r e m 6 . 4 . A s s u m e (1). T o prove (2), suppose that A C \ A > 1 for some element x G. W e work to show t h a t x G A . Since G = A N , we can w r i t e x = a n w i t h a G A and n G N , and thus A = A = A . T h e n A n A > 1, and so we can choose a n o n i d e n t i t y element b of this intersection, where b G A and b G A . W e thus have [ b , n ] = b ~ b G A , and since N < G, we also have [b, n] G N . T h e n [b, n] G A n / V = 1, a n d hence 6 centralizes n . B u t the action of A o n N is Frobenius by (1) and 6 ^ 1 , and it follows t h a t n = l . T h u s x = a n = a lies i n A , as required.
x a n n n n n x n

N o w assume (2), and let 1 ^ a G A . If x G C ( A ) , t h e n a G A n A , a n d since this intersection is n o n t r i v i a l , it follows b y (2) t h a t x G A T h u s C ( o ) C A , p r o v i n g (3).
x G G

N e x t , we show that (3) (1). If 1 ^ a G A , then by (3), we have CJV(O ) = TV n C ( o ) C / V n A = 1, a n d thus the action of A on TV is Frobenius, w h i c h is assertion (1). W e have now established that (1), (2) and (3) are equivalent.
G

A s s u m i n g (1) and (2) now, we and C ( n ) 2 N , then n = 1. B y every element of G outside of N lies t h a t some nonidentity element of C
G

prove (4) by showing t h a t if n G N (2) and C o r o l l a r y 6.6, we know t h a t i n some conjugate of A , and it follows ( n ) lies i n a conjugate of A . F o r an

184

6. F r o b e n i u s

Actions

appropriate conjugate m of n , therefore, some n o n i d e n t i t y element of C ( m ) lies i n A . Since m G N a n d the action of A on N is assumed to be Frobenius, it follows t h a t m = 1, and thus n = 1, p r o v i n g (4).
G

F i n a l l y , assuming (4), let 1 + n G N . T h e n C ( n ) = A D C ( n ) C A i l J V = l , and thus the a c t i o n of A on N is Frobenius, p r o v i n g (1).

A G

In fact, the overall assumption i n T h e o r e m 6.4 t h a t G has a n o r m a l subgroup N complemented b y a subgroup A is not entirely necessary. B y F r o b e n i u s ' theorem, c o n d i t i o n (2), w h i c h concerns the embedding of A i n G, guarantees the existence of the n o r m a l subgroup N for w h i c h A is a complement. Somewhat analogously, c o n d i t i o n (4), w h i c h concerns the embedding of N i n G, guarantees the existence of the complement A for N . T h i s result, however, is far more elementary t h a n Frobenius' theorem. 6.7. T h e o r e m . L e t N < G, w h e r e G i s a finite g r o u p , a n d suppose C (n) C N f o r every n o n i d e n t i t y element n G N . T h e n N i s c o m p l e m e n t e d i n G, a n d i f K N < G, t h e n G i s a F r o b e n i u s g r o u p w i t h k e r n e l N .
G

P r o o f . L e t p be a p r i m e divisor of \ N \ . Choose P G S y \ ( N ) and S G Sylp(G), with P C S . T h e n Z ( S ) C C ( P ) , and since P is n o n t r i v i a l and is contained i n N , it follows by hypothesis that C ( P ) C N , and thus Z ( S ) C N . N o w S is n o n t r i v i a l and hence Z ( S ) is a n o n t r i v i a l subgroup of N . B y hypothesis, therefore, S C C ( Z ( 5 ) ) C N . T h u s N contains a full Sylow p-subgroup of G for every prime p d i v i d i n g \ N \ , and it follows t h a t no such p r i m e divides \ G : N \ . T h i s shows t h a t TV is a n o r m a l H a l l subgroup of G, and hence it has a complement A i n G b y the Schur-Zassenhaus theorem. If 1 < N < G then b o t h N and A are n o n t r i v i a l , and so G is a Frobenius group b y T h e o r e m 6.4.
p G G G

Problems 6A 6 A . 1 . L e t A be a subgroup of S L ( 2 , p ) of order not divisible by the prime p . Show that the action of A on the vector space of order p is Frobenius.
2

6 A . 2 . I n the m u l t i p l i c a t i v e group F

of a field F of order 43, let a have

order 7 and let e have order 3. W o r k i n g i n the group G L ( 3 , 4 3 ) , let a 0 0 0 a 0 0 a 0 0 1 0 0 0 1 0

a =

and

Show t h a t A = ( a , b) is noncyclic of order 63, and that its n a t u r a l action on the vector space V of order 4 3 is Frobenius.
3

Problems 6 A

185

H i n t . C h e c k t h a t (b) has order 9 and normalizes (a), w h i c h has order 7. A l s o , observe t h a t to show t h a t the a c t i o n is Frobenius, it suffices to check t h a t C { t ) = 1 for elements i e A o f prime order.
v

N o t e . T h i s p r o b l e m shows t h a t a nonabelian group of o d d order can be a Frobenius complement. 6A.3. Show t h a t there is a noncyclic group of order 5 - l l t h a t has a Frobenius a c t i o n on some n o n t r i v i a l group. H i n t . W o r k i n G L { 5 , p ) for some appropriate p r i m e p . 6A.4. L e t G be a Frobenius group w i t h Frobenius kernel N . Show t h a t each coset of N i n G other t h a n N itself is contained i n a single conjugacy class of G. 6A.5. L e t G be a nonabelian solvable group i n w h i c h the centralizer of every nonidentity element is abelian. Show t h a t G is a Frobenius group, where F ( G ) is the Frobenius kernel. 6A.6. L e t A > 1 and B > 1 satisfy the hypothesis of L e m m a 6.5 i n G. Show t h a t A n B > 1 for some element g G G.
9 2

H i n t . Consider X and Y corresponding to A and B as i n L e m m a 6.5. N o t e . Since we have not proved F r o b e n i u s ' theorem t h a t the set appearing i n L e m m a 6.5 is a c t u a l l y a subgroup, y o u should not appeal to t h a t fact i n these problems. 6A.7. H L e t A satisfy the hypotheses of L e m m a 6.5 i n G, a n d assume A C (a) G i v e n g G G, show t h a t if A (b) If H < G, show t h a t H = G. H i n t . F o r (a), check t h a t A n H satisfies the hypothesis of L e m m a 6.5 i n H and use the previous problem. 6A.8. L e t G, A and X be as i n L e m m a 6.5. If M < G, show t h a t either M C X or X C M . H i n t . If M n A > 1, use the previous p r o b l e m to show t h a t A M = G. 6A.9. L e t G, A and A be as i n L e m m a 6.5, a n d suppose t h a t t G A is a n involution. (a) Show t h a t there are at least \ G : A \ - 1 elements i e G - i t h a t re* = x - K such
9

C G.
9

n H > 1, t h e n g G H .

186

6. F r o b e n i u s

Actions

(b) Show t h a t t inverts every element of the set X . (c) Show t h a t t is the unique i n v o l u t i o n i n A , and i n particular, t is central i n A . (d) Show t h a t every element of X has o d d order. (e) Suppose that x and y lie i n X and that xy~ x = y and deduce that x = y .
2 2 l

A . Show that

(f)

Show t h a t X is a subgroup.

H i n t . L e m m a 2.14(b) is relevant for (a). N o t e . T h i s p r o b l e m provides a proof of Frobenius' theorem i n the case that the subgroup A has even order. 6 A . 1 0 . L e t G , A and X be as i n L e m m a 6.5. (a) Show for each p r i m e divisor p of \ A \ that A contains a full Sylow p-subgroup P of G and t h a t A controls G-fusion i n P . (b) If A > 1, show t h a t G ' A = G and G ' n A = A ' . (c) If A is solvable, show t h a t the subset X is a subgroup. N o t e . B y this p r o b l e m and the previous one, we know t h a t the conclusion of F r o b e n i u s ' theorem holds i f either \ A \ is even or A is solvable. B y the FeitT h o m p s o n odd-order theorem, one of these possibilities must necessarily occur, a n d this shows t h a t Frobenius' theorem is a consequence of the o d d order theorem. B u t since the F e i t - T h o m p s o n proof uses character theory heavily, this certainly does not y i e l d a character-free proof of Frobenius' theorem. 6 A . 1 1 . L e t A C G . Show t h a t A satisfies the hypothesis of L e m m a 6.5 i n G if a n d o n l y if N ( T ) C A for every nonidentity subgroup T of A .
G

6B In this section we present necessary conditions for a group A to be a Frobenius complement. (Recall that we have defined this to mean t h a t there exists some nonidentity group N a n d a Frobenius action of A on N . ) T h e key idea is to consider "partitioned" groups. A p a r t i t i o n of G is a collection II of nonidentity proper subgroups of G such that | J II = G and X n Y = 1 for every two distinct members X a n d Y of II. F o r example, if G = D , the dihedral group of order 2n w i t h n > 1, then G is p a r t i t i o n e d by the set II consisting of the cyclic subgroup C of order n and the n subgroups of order 2 t h a t are not contained i n C.
2 n

An

elementary abelian p-group G of order p

w i t h e > 1 provides an-

other example of a p a r t i t i o n e d group.

In this situation, the set II of a l l

6B

187

subgroups of order p i n G is a p a r t i t i o n . It is clear t h a t every two distinct members of H" intersect t r i v i a l l y , a n d also, since every n o n i d e n t i t y element of G has order p , each such element lies i n some subgroup of order p , and so U n = G, as required. In what follows, the c a r d i n a l i t y of a p a r t i t i o n w i l l be relevant, and i n this case, it is especially easy to compute. E a c h member of n contains exactly p - 1 nonidentity elements, a n d since G has a t o t a l of p - 1 n o n i d e n t i t y elements, we see t h a t II consists of exactly ( p - 1)/(p - 1) subgroups. I n p a r t i c u l a r , if e = 2, t h e n |II| = p + 1.
e e

We m e n t i o n one more family of examples. If G is a Frobenius group w i t h kernel N and complement A , let II be the collection of subgroups consisting of N and a l l conjugates of A . Since N n A = 1, it is clear that N intersects t r i v i a l l y w i t h every conjugate of A . A l s o , every two distinct conjugates of A intersect t r i v i a l l y because A (and each of its conjugates) satisfies statement (2) of T h e o r e m 6.4. Furthermore, b y C o r o l l a r y 6.6, we know t h a t every element of G not i n N lies i n some conjugate of A , and thus [j U = G. In this s i t u a t i o n , A = N ( A ) , and so the number of conjugates of A i n G is \ G : A \ = \ N \ , and it follows t h a t |II| = 1 + \ N \ .
G

6 . 8 . L e m m a . L e t U be a p a r t i t i o n of a finite g r o u p A , a n d suppose that A acts v i a a u t o m o r p h i s m s o n a n a b e l i a n g r o u p U . I f U has a n element with o r d e r n o t d i v i d i n g |XT| - I , t h e n t h e r e exists a member X G II such that C u { X ) > 1. P r o o f . F o r each subgroup H C A and each element u G U, write

and observe t h a t this element of U is u n a m b i g u o u s l y defined since the order of the factors i n the product is irrelevant because U is abelian. A l s o , since the a c t i o n b y an element h H s i m p l y permutes the factors of u , it follows that ( u ) = u for a l l h e H , and thus u G C { H ) . Since we can assume t h a t C u ( X ) = 1 for each member X G II, we have u = 1 for a l l u G U and all X G n .
H h H H H V x

Now

fix u G U, and compute t h a t

xen where the second equality follows because as a runs over a l l of the elements of a l l of the subgroups X G II, each n o n i d e n t i t y element of A occurs e x a c t l y once, while the identity of A occurs a t o t a l of )H| times. If we choose u G U such t h a t u l ! " ^ 1, it follows t h a t u + 1, a n d since u G C (A) C C r j ( X ) for a l l X G II, the proof is complete.
1 1 1 A A V

188

6. F r o b e n i u s A c t i o n s

6 . 9 . T h e o r e m . L e t A be a F r o b e n i u s c o m p l e m e n t . T h e n n o s u b g r o u p of A i s a n e l e m e n t a r y a b e l i a n p - g r o u p of o r d e r e x c e e d i n g p , a n d n o s o l v a b l e s u b g r o u p of A i s a F r o b e n i u s g r o u p . A l s o , every s u b g r o u p of A h a v i n g o r d e r p q , w h e r e p a n d q are (possibly equal) p r i m e s i s cyclic. P r o o f . L e t A act on N , where the a c t i o n is Frobenius a n d N > 1. Suppose t h a t A has a subgroup t h a t is either an elementary abelian p-group of order p > p or is a solvable Frobenius group. Since the a c t i o n of every subgroup of A o n N is Frobenius, we can replace A b y the appropriate subgroup and assume t h a t A itself is either elementary abelian of order p or is solvable and Frobenius. A l s o , i n the elementary abelian case, we can replace A by a subgroup if necessary, and so we can assume t h a t e = 2.
e e

In either case, A is solvable and acts coprimely on N , and so i f we choose any p r i m e divisor r of \ N \ , T h e o r e m 3.23 guarantees the existence of a n A invariant Sylow r-subgroup R of N . T h e n R > 1, and so Z ( R ) > 1, and since Z ( R ) is characteristic i n R, this subgroup is A - i n v a r i a n t . Since the action of A on Z ( R ) is Frobenius, we c a n replace N by Z ( R ) , and hence we can assume t h a n N is abelian. Since A is either elementary abelian of order p or is a Frobenius group, A has a p a r t i t i o n II of c a r d i n a l i t y 1 + n , where n is a divisor of \ A \ . (If A is elementary of order p , then n = p , a n d if A is Frobenius, then n is the order of its Frobenius kernel.) Since n divides \ A \ , it is coprime to | N | , and thus since N is n o n t r i v i a l , some element e J V satisfies u ^ 1. (In fact, we can take u to be any n o n i d e n t i t y element of N . ) It follows b y L e m m a 6.8 t h a t C ( X ) > 1 for some member X G II. B u t since \X\ > 1, this contradicts the fact t h a t the a c t i o n of A on N is Frobenius, a n d we conclude that as claimed, the original group A cannot c o n t a i n an elementary abelian p-group of order exceeding p or a solvable Frobenius group.
2 n N 2

W h a t remains is to show t h a t if B C A , where \ B \ = p q , and p and q are possibly equal primes, t h e n B is cyclic. F i r s t , if p = q, then \ B \ = p , and since we have shown that B eannot be elementary abelian, the only other possibility is t h a t it is cyclic, as required. F i n a l l y , assume t h a t p > q. T h e n B must have a n o r m a l Sylow p-subgroup P of order p. If Q G S y l , ( B ) and Q acts (by conjugation) n o n t r i v i a l l y on P , we have seen t h a t the action of Q on P must be Frobenius. T h u s B is a solvable Frobenius group, w h i c h we k n o w is impossible. It follows t h a t Q centralizes P , a n d it is easy to see in this case t h a t B = P Q is cyclic. T h i s completes the proof.
2

6.10. C o r o l l a r y . L e t P G Sylp(A), w h e r e A i s a F r o b e n i u s T h e n P c o n t a i n s a t most one s u b g r o u p of o r d e r p .

complement.

P r o o f . W e can assume t h a t P > 1. T h e n Z ( P ) > 1, and we can choose Z C Z { P ) w i t h \ Z \ = p . W e argue t h a t Z is the only subgroup of order p

6B

189

in P . Otherwise, let U C P be a second subgroup of order p, a n d observe t h a t Z U is a subgroup of order p since Z < P . A l s o , Z P is not cyclic since it contains two subgroups of order p , and this contradicts T h e o r e m 6.9.
2

We digress now to classify the p-groups t h a t have at most one subgroup of order p, as i n C o r o l l a r y 6.10. Some obvious examples are the cyclic pgroups, a n d if p = 2, also the generalized q u a t e r n i o n 2-groups. These, we recall, are the 2-groups P of order at least 8 t h a t have a cyclic subgroup C = (c) of index 2 and a n element a of order 4 i n P - C such t h a t c = c~\ It is not h a r d to see t h a t if P is generalized quaternion and C is cyclic of index 2 as above, then every element of P - C has order 4. T h e subgroup of order 2 i n C, therefore, is the unique subgroup of order 2 i n P .
a

In fact, cyclic p-groups and generalized quaternion 2-groups are the only p-groups t h a t have just one subgroup of order p. 6.11. T h e o r e m . L e t P be a p - g r o u p c o n t a i n i n g a t most one s u b g r o u p of o r d e r p . T h e n e i t h e r P i s c y c l i c , o r else p = 2 a n d P i s g e n e r a l i z e d q u a t e r nion. W i t h only a little e x t r a work, we can prove a stronger theorem, and so a l t h o u g h the next result is not directly relevant to Frobenius actions, it seems reasonable to present it here. W e shall see t h a t T h e o r e m 6.11 is an i m m e d i a t e consequence. It is easy to see that a n a b e l i a n p-group w i t h at most one subgroup of order p must be cyclic. ( T h i s is i m m e d i a t e from the fundamental theorem of abelian groups.) It follows t h a t i f P is any p-group t h a t has at most one subgroup of order p, t h e n a l l abelian subgroups of P are cyclic. In fact, if p ^ 2, the apparently m u c h weaker a s s u m p t i o n t h a t a l l n o r m a l abelian subgroups of P are cyclic is sufficient to guarantee that P is cyclic. F o r p = 2, however, there are some possibilities other t h a n cyclic and generalized quaternion groups. 6.12. T h e o r e m . L e t P be a p - g r o u p i n w h i c h every n o r m a l a b e l i a n s u b g r o u p is c y c l i c . T h e n e i t h e r P i s c y c l i c , o r e l s e p = 2 a n d P i s d i h e d r a l , g e n e r a l i z e d quaternion o r semidihedral. R e c a l l t h a t a 2-group P is dihedral if it has a cyclic subgroup C = ( c ) of index 2 and an involution a e P - C such t h a t c = c~\ Like dihedral and generalized quaternion 2-groups, a semidihedral 2-group also has a cyclic subgroup of index 2. Specifically, a 2-group P of order at least 16 is semidihedral if it has a cyclic subgroup C = (c) of i n d e x 2, and there is a n i n v o l u t i o n a i n P - C such t h a t c = zc~ , where z is the unique element of order 2 i n C. (Note t h a t we need to assume t h a t \ P \ > 16 because i f | P | = 8, t h e n c has order 4, and so zc~ = c. I n t h a t case, a w o u l d centralize C and P
a a 1 l

190

6. F r o b e n i u s

Actions

w o u l d be abelian.) It is not h a r d to see t h a t most d i h e d r a l 2-groups, as well as a l l generalized quaternion and semidihedral 2-groups a c t u a l l y do satisfy the hypotheses of T h e o r e m 6.12. (The exception among d i h e d r a l groups is D , of order 8, w h i c h has two noncyclic n o r m a l abelian subgroups of order
8

4-) In a d i h e d r a l or semidihedral group, there is by definition at least one i n v o l u t i o n outside of the cyclic subgroup of index 2. These groups, therefore, have more t h a n one subgroup of order 2, and so they do not satisfy the hypothesis of T h e o r e m 6.11. It follows that T h e o r e m 6.11 is a consequence of T h e o r e m 6.12. We begin w o r k t o w a r d a proof of T h e o r e m 6.12 w i t h the following.

6.13. L e m m a . L e t P be a n o n a b e l i a n 2 - g r o u p h a v i n g a c y c l i c s u b g r o u p C = (c) of i n d e x 2, a n d l e t a G P - C. If c = c~ , t h e n P i s d i h e d r a l o r g e n e r a l i z e d q u a t e r n i o n , a n d if c = zc~ , w h e r e z i s t h e u n i q u e i n v o l u t i o n i n C, t h e n P i s s e m i d i h e d r a l .

a l a l

P r o o f . Since P is nonabelian and P = (a,c), the elements a and c cannot commute. I n the case where c = c~\ it follows t h a t the order o(c) > 4 and i n the case where c = zc~\ we have o(c) > 8. N o w i f c = c T , then a inverts every element of C, and if c = zc' , then ( c ) = ( c ) " , and so a inverts every nongenerating element of C. I n b o t h cases, therefore, every element of order 4 i n C is inverted by a . Since a G C is centralized b y a, the group (a ) can contain no element of order 4, a n d thus either o ( a ) = 2 or a = z and o ( a ) = 4. If c = c " , therefore, P is either dihedral or generalized quaternion, respectively, and if c = zc~ a n d o ( a ) = 2, then P is semidihedral. I n the r e m a i n i n g case, a = z and c = zc' , and thus ( c a ) = c a c a = c a c = 1. T h e n o(ca) = 2 and since = c , we can replace a by c a to see t h a t P is semidihedral.
a a a 1 a 1 2 a 2 1 2 2 2 a 1 a l 2 a 1 2 2 a a

A n easy consequence is the following description of the nonabelian groups of order 8, w h i c h we w i l l need i n what follows. 6 . 1 4 . C o r o l l a r y . L e t P be a n o n a b e l i a n g r o u p of o r d e r 8. T h e n P i s i s o m o r p h i c either t o the dihedral group D o rt o the q u a t e r n i o n group Q.
8 8

P r o o f . Since P is nonabelian, not every nonidentity element is a n involut i o n . W e can thus choose c G P of order 4, and we w r i t e C = (c), so that C has index 2 i n P . If a G P - C then since P is nonabelian, a cannot centralize c, and so a induces a n o n t r i v i a l a u t o m o r p h i s m of C. T h e cyclic group C of order 4, however, has only one n o n t r i v i a l a u t o m o r p h i s m , and we deduce t h a t c = c' . T h e assertion now follows by L e m m a 6.13.
a 1

6B

191

6.15. L e m m a . L e t T be a p - g r o u p w i t h o r d e r d i f f e r e n t f r o m 8, a n d suppose t h a t \ T : Z ( T ) | = p . L e t C be a c y c l i c s u b g r o u p such t h a t Z ( T ) < C < T . Then T has a c h a r a c t e r i s t i c e l e m e n t a r y a b e l i a n s u b g r o u p of o r d e r p .

2 2

P r o o f . Since C < T is abelian and T / C is cyclic, we c a n a p p l y L e m m a 4.6 to conclude t h a t \ C \ = | T ' | | Z ( T ) | . It follows t h a t \T'\ = p , and thus T' C Z ( T ) , and T has nilpotence class 2. F i r s t , suppose t h a t p + 2. B y T h e o r e m 4.8, the m a p 9 : T - > T denned b y 9 { x ) = x ? is a h o m o m o r p h i s m , and we let K = ker (0). T h e n K = { x G P | x P = 1} is a characteristic subgroup of T, and it suffices to show t h a t \ K \ = p . Since T / Z { T ) cannot be cyclic since T is nonabelian, this factor group of order p must be elementary abelian. T h u s 9 ( T ) C Z(T), and it follows t h a t \ K \ = \ T \ / \ 9 { T ) \ > | T | / | Z ( T ) | = p . F i n a l l y , \ K : K n C\ < \ T : C\ = p , and since K n C is a cyclic subgroup of i f , we have | i f n C\ < p . It follows t h a t \ K \ < p , and we are done i n this case.
2 2 2 2

We can now assume t h a t p = 2. I n w h a t follows, we w i l l twice use the easy observation t h a t if A is a noncyclic abelian 2-group t h a t has a cyclic subgroup of index 2, then { a G A | a = 1} is a characteristic element a r y abelian subgroup of order 4. F i r s t , a p p l y i n g this to the group T/T' (which is not cyclic because T' is central a n d T is not abelian) we o b t a i n a characteristic elementary abelian subgroup E/T' of order 4, and thus E is characteristic i n T and | | = 8.
2

Now E is noncyclic, and hence E ^ C . A l s o , ET since | T | / 8, and we conclude t h a t E 2 C. It follows t h a t n C is a proper subgroup of C, w h i c h , therefore, is contained i n Z { T ) . A l s o , E n C i s a cyclic subgroup of E w i t h index 2, and hence E is abelian but not cyclic. It follows t h a t E has a characteristic elementary abelian subgroup AT of order 4, and since is characteristic i n T, so too is K . T h i s completes the proof.

N e x t , we present an elementary number-theoretic fact. 6.16. L e m m a . L e t p be p r i m e , a n d l e t e be a p o s i t i v e i n t e g e r . Suppose that i i s a n i n t e g e r such t h a t i EE 1 m o d p . Then one of t h e f o l l o w i n g o c c u r s .
p e

(a) i EE 1

modp*- .

(b) p = 2 and

i ==-1 m o d 2 .
e

(c) p = 2 and i

EE

2 "

- 1 mod 2 .

P r o o f . F i r s t , suppose t h a t p > 2. If i = 1, t h e n (a) certainly holds, a n d so we c a n assume t h a t i + 1, and we write i - l = m p where p does not d i v i d e m a n d i > 0. W e have i = i = 1 m o d p , where the first congruence follows
t p

192

6. F r o b e n i u s A c t i o n s

by Fermat's little theorem, and thus t > 0. A l s o , i

p

= (1 + mp'y

EE 1 + m p

+ (fyrn p

2 t

mod p

Since p is o d d , the b i n o m i a l coefficient (P) is divisible by p , and we have iP EE 1 + m p mod p . It follows t h a t the highest power of p d i v i d i n g %P - 1 is p . B y hypothesis, however, p divides *P - 1, and thus e < t + l . T h e n e - 1 < t, and we have i = 1 m o d p ~ , as required.
t + 1 2 i + 1 t + 1 e e l

Now assume p = 2, so t h a t 2 divides i - 1 = ( i +1). Since one of the factors i - 1 or i + 1 is not divisible by 4, the other must be divisible by 2 " , and thus i = 1 m o d 2 " . A s s u m i n g t h a t (a) fails, we have i = - 1 mod 2 ~ \ and we can w r i t e i = - 1 + a 2 " for some integer a . If a is even, t h e n i EE 1 m o d 2 and (b) holds. Otherwise, a = 2 b + 1 for some integer 6, and so i = - 1 + (26 + l ) 2 - EE 2 ~ - 1 m o d 2 , p r o v i n g (c).
6 1 6 1 e e 1 e e 1 e l e

We

can now assemble the pieces.

P r o o f o f T h e o r e m 6 . 1 2 . L e t C be m a x i m a l a m o n g abelian n o r m a l subgroups of P . B y hypothesis, C is cyclic, so we can assume that C < P . A l s o , C = C ( C ) since otherwise, we could find a subgroup B < P w i t h C C B C C p ( C ) , where | B : C | = p. Since C is central i n B and B / C is cyclic, it w o u l d follow t h a t B is abelian, w h i c h contradicts the choice of C. T h u s P / C = P / C { C ) is i s o m o r p h i c a l l y embedded i n A u t ( C ) , w h i c h is abelian, a n d thus P / C is abelian. A l s o , if \ C \ = 4, it follows t h a t \ P / C \ = 2 and \ P \ = 8, and by C o r o l l a r y 6.14, there is n o t h i n g to prove i n this case. We can assume, therefore, that \ C \ + 4, and we write \ C \ = p .
P P e

Let T / C C P / C have order p , and observe that T < P since P / C is abelian. A l s o , T is not abelian and \T\ / 8 since \ C \ / 4. Since T cannot have a characteristic elementary abelian subgroup of order p , it follows by L e m m a 6.15 t h a t P cannot be central i n T , where C = (c). T h u s ( P ) ^ where we have chosen a e T - C .
2 C C a

We

can w r i t e c = c for some integer i. Since aP G C , it follows t h a t

a i

c = c
e

= c

,
e

a n d since c has order p , we conclude that %P EE 1 m o d p . Furthermore, P ^ ( P ) = ( c P f , a n d since c has order p ~ \ we conclude t h a t i EE 1 m o d p " . It follows by L e m m a 6.16 t h a t p = 2 a n d i = - 1 m o d 2 " . T h u s
a p e C c 6 1 6 1

(c r
2

= (c )- .

Now P / C acts faithfully o n C , a n d b y the c o m p u t a t i o n of the previous paragraph, every i n v o l u t i o n i n the abelian group P / C acts to invert the element c , w h i c h has order at least 4. It follows t h a t there is o n l y one such i n v o l u t i o n because if there were two, their product w o u l d be a n i n v o l u t i o n
2

6B

193

that centralizes and does not invert c . Since P / C is an abelian 2-group w i t h just one involution, we conclude t h a t P / C is cyclic. We argue next t h a t \ P / C \ = 2. Otherwise, the element a is a square m o d u l o C, and it follows that i = j m o d 2 for some integer j . B u t i = - 1 mod 2 " \ and since e > 3, it follows t h a t j = - 1 m o d 4, and this is not possible.
2 e e 2

Now mod
e

P = T, and by L e m m a 6.16, either i = - 1 m o d 2 or i = 2 ~

a

- 1
1

2 . I n other words, either c

= c~

or c

= zc~\

where z = c " "

is

the i n v o l u t i o n i n C. T h e result now follows by L e m m a 6.13. We r e t u r n now to the study of Frobenius complements.

6.17. C o r o l l a r y . Suppose that A is a Frobenius complement. S y l o w s u b g r o u p of A i s c y c l i c o r g e n e r a l i z e d q u a t e r n i o n .

Then

each

P r o o f . B y C o r o l l a r y 6.10, a Sylow p-subgroup of A has at most one subgroup of order p . B y T h e o r e m 6.11, such a group must be cyclic or generalized quaternion. Observe t h a t by C o r o l l a r y 6.17, a l l Sylow subgroups of a n abelian Frobenius complement A are cyclic, and so A itself must be cyclic. 6.18. C o r o l l a r y . L e t A be a F r o b e n i u s c o m p l e m e n t and A / A ! are cyclic and have coprime orders. of o d d o r d e r . Then A'

P r o o f . B y C o r o l l a r y 6.17, a l l Sylow subgroups of A are cyclic. T h e result now follows by T h e o r e m 5.16.

We k n o w more about the structure of a F r o b e n i u s complement t h a n what its Sylow subgroups look like, a n d we c a n use this to derive more subtle i n f o r m a t i o n . F o r example, we have the following. 6.19. T h e o r e m . L e t A be a F r o b e n i u s c o m p l e m e n t of o d d o r d e r . has a u n i q u e s u b g r o u p of o r d e r p f o r each p r i m e p d i v i d i n g \ A \ . Then A

P r o o f . F i r s t , consider a prime divisor r of \ A ' \ . T h e n r does not d i v i d e \ A / A ' \ by C o r o l l a r y 6.18, and thus every subgroup R of order r i n A a c t u a l l y lies i n A ' . Since A ' is cyclic, however, it has only one subgroup of order r , and thus R is unique, and i n p a r t i c u l a r , R < A . Now let q be a p r i m e divisor of \ A \ such t h a t q does not divide \ A ' \ . T h e n q divides the order of the cyclic group A / A ' , a n d so A / A ' has a unique subgroup X / A ' of order q. T h e n every subgroup Q C A of order q lies i n X , and i n fact X = A ' Q . A l s o , Q G S y l ( A ) since q does not divide \ A ' \ . T o prove t h a t Q is unique, therefore, it suffices to show t h a t Q < X .
g

194

6. F r o b e n i u s

Actions

Let R C A ' be an a r b i t r a r y subgroup of p r i m e order, say r . W e know t h a t R < A , a n d thus R Q is a subgroup of A of order r q . It follows that R Q is cyclic by T h e o r e m 6.9, and i n particular, Q centralizes R. T h u s C > ( Q ) contains every subgroup of A ' having prime order.
A

Since A ! is abelian and has order coprime to q, it follows by F i t t i n g ' s theorem t h a t A ' = [ A ' , Q] x C . ( Q ) . Since the second factor contains every p r i m e order subgroup of A ' , the first factor can contain no such subgroup, and we conclude t h a t [ A ' , Q ] = 1. T h u s Q centralizes A ' , and so Q < A ' Q = X , as required.
A

N e x t , we present two results t h a t are needed i n C h a p t e r 7. 6 . 2 0 . L e m m a . Suppose t h a t A i s a finite a b e l i a n g r o u p t h a t acts faithfully o n a finite g r o u p N , w h e r e \ A \ a n d \ N \ a r e c o p r i m e . Assume i n addition t h a t t h e a c t i o n of A o n every p r o p e r A - i n v a r i a n t s u b g r o u p of N i s t r i v i a l . T h e n A is cyclic. We 6.21. on a w i l l deduce this as a corollary of the following. T h e o r e m . Suppose finite that A is a finite abelian group and that i t acts

g r o u p N , w h e r e \ A \ a n d \ N \ a r e c o p r i m e . If A i s n o t c y c l i c , t h e n N = { C ( a ) | 1 / a A ).
N

P r o o f . T h e assertion is t r i v i a l i f N = 1, so we assume t h a t TV > 1, and we proceed by i n d u c t i o n o n \ N \ . W r i t e K = { C ( a ) | 1 / a A ),

N

so t h a t our goal is to show that K = N . If M is a proper subgroup of N that admits the action of A , then by the inductive hypothesis, M = (C {a) | 1/aG
M

A ) CAT,

and so K contains every proper A - i n v a r i a n t subgroup of N . N o w assume t h a t K < N , and choose a p r i m e divisor p of \ N : K \ . Since A is certainly solvable, a n d its order is coprime to \ N \ , we know that there exists some A invariant Sylow p-subgroup P of N , and we see that P % K since p divides |TV : K \ . It follows t h a t P = N , and so TV is a p-group, and thus N ' is a proper A - i n v a r i a n t subgroup. T h e n N ' C K , and it follows that K < N . A l s o , K is A - i n v a r i a n t since it is uniquely determined by A and N . Now for consider the induced action of A on N = N / K . Since "fixed points C (a)
N

come from fixed points" i n coprime actions, it follows t h a t C ^ ( a ) = a l l elements a < E A . B u t if 1 ^ a < E A , then C ( a )
N w

C K , and so

C (a)
N

is t r i v i a l . It follows t h a t C { a ) is t r i v i a l , and thus the action of A on the

Problems 6 B

195

n o n t r i v i a l group N is Frobenius. B u t A is not a Frobenius complement since it is abelian but not cyclic. T h i s c o n t r a d i c t i o n completes the proof. P r o o f o f L e m m a 6 . 2 0 . Since A is abelian, the subgroup C ( a ) is A invariant for a l l a G A . ( T h i s is because C { a ) is u n i q u e l y determined b y the element a, w h i c h is invariant under conjugation by A . ) A l s o , if a + 1, then C ( a ) < N because the a c t i o n of A o n N is faithful, and thus by hypothesis, C ( a ) C C { A ) for a l l n o n i d e n t i t y elements a G A . If A is not cyclic, however, then by T h e o r e m 6.21, the subgroups C { a ) generate N , a n d thus N = C ( A ) . Since A acts faithfully, this yields A = 1, and this is a contradiction, since we assumed t h a t A is not cyclic. T h e result now follows.
N N N N N N N

Problems

6B

6 B . 1 . Show t h a t every Frobenius group contains a solvable Frobenius subgroup, and deduce t h a t a Frobenius complement cannot have a Frobenius group as a subgroup.

6 B . 2 . L e t A act faithfully on N , where A is abelian, a n d assume t h a t no nonidentity proper subgroup of N is A invariant. Show t h a t A is cyclic. H i n t . L e t p be a prime divisor of \ N \ and let P G S y l ( A ) . T h e n C ( P ) A - i n v a r i a n t and n o n t r i v i a l .
p N

is

6 B . 3 . Show t h a t if the coprimeness assumption i n T h e o r e m 6.21 is dropped, the result becomes false.

6 B . 4 . Suppose t h a t G has a p a r t i t i o n IT such that [ X , Y] = 1 for every two distinct members X and Y of IT. (a) Show t h a t G is abelian. (b) Show that the elements of G have equal prime orders, so that G is elementary abelian. H i n t . F o r (b), consider x G X G IT and y G Y G IT, where X / Y. If o { x ) < o ( y ) , consider the element { x y ) - \ I n w h i c h member of IT does it lie?
o l x

N o t e . T h e hypotheses of the previous p r o b l e m are a u t o m a t i c a l l y satisfied if the members of IT are a l l n o r m a l i n G.

196

6. F r o b e n i u s

Actions

6 B . 5 . Suppose t h a t G has a p a r t i t i o n consisting of s u b n o r m a l subgroups. Show t h a t G is nilpotent. H i n t . W o r k i n g by i n d u c t i o n on \ G \ , show that every subgroup H < G t h a t is not contained i n some member of the p a r t i t i o n is nilpotent. Deduce that every member of the p a r t i t i o n that is not contained i n F ( G ) is n o r m a l i n G a n d has p r i m e index. 6 B . 6 . L e t C be a cyclic 2-group of order at least 8. Show t h a t C has exactly three automorphisms of order 2. 6 B . 7 . L e t P be a nonabelian 2-group t h a t has a cyclic subgroup of index 2. If \ P : Z ( P ) | > 4, show that P is dihedral, semidihedral or generalized quaternion. 6 B . 8 . L e t P be a 2-group of order at least 8, and assume t h a t \ P : P ' \ = 4. Show t h a t P is dihedral, semidihedral or generalized quaternion. H i n t . L e t Z C G ' n Z ( G ) have order 2. W o r k i n g by i n d u c t i o n on deduce t h a t P has an abelian subgroup A of index 2 and t h a t A is either cyclic or is the direct product of Z w i t h a cyclic group. In the latter case, and assuming t h a t | P | > 16, deduce t h a t Z < Z ( P ) and derive a c o n t r a d i c t i o n . N o t e . T h i s p r o b l e m proves a theorem of O . T a u s s k y - T o d d . 6 B . 9 . L e t G be solvable, and assume t h a t every element of G has primepower order. Show t h a t | G | is divisible by at most two primes. 6C We now begin work toward a proof of the theorem of T h o m p s o n ' s P h . D . thesis: Frobenius kernels are nilpotent. W e prove this first under the addit i o n a l assumption t h a t the given Frobenius kernel is solvable. ( T h i s special case, w h i c h is needed for T h o m p s o n ' s result, appears i n a 1957 paper of G . H i g m a n , where it is described as being already known.) 6.22. T h e o r e m . L e t N be a s o l v a b l e F r o b e n i u s k e r n e l . Then N is nilpotent.

P r o o f . B y hypothesis, there exists a n o n t r i v i a l group A h a v i n g a Frobenius action on N . Since the action of every subgroup of A on N is also Frobenius, we c a n replace A by a subgroup of prime order, and so we c a n assume that \ A \ = p , & prime. L e t G = N x > A , and as usual, view N and A as subgroups of G , where N < G is complemented by A . A s s u m i n g t h a t N is not nilpotent, we proceed by i n d u c t i o n on |7V|. Since N > 1, we can choose a m i n i m a l n o r m a l subgroup U of G, w i t h U C N . The i n d u c e d action of A o n N / U is Frobenius by T h e o r e m 6.2, and of course

6C

197

N / U is solvable. B y the inductive hypothesis, therefore, N / U is nilpotent, and thus 7V C U. B u t N > 1 since N is not nilpotent, and of course, 7V < G. T h e m i n i m a l i t y of U, therefore, yields U = N, and thus U is the unique m i n i m a l n o r m a l subgroup of G contained i n N . ( A n y other one w o u l d also have to be N.) Since U is solvable and is m i n i m a l n o r m a l i n G, L e m m a 3.11 guarantees t h a t U must be an abelian g-group for some p r i m e q. B u t N is not a g-group since it is not nilpotent, a n d so we can choose a p r i m e divisor r / q of \ N \ . Let R e S y \ ( N ) , a n d choose R to be A - i n v a r i a n t . ( A l t h o u g h the existence of an A - i n v a r i a n t Sylow r-subgroup follows v i a T h e o r e m 3.23, we do not really need the full strength of that theorem here. T h i s is because | A | = p a n d |SyL.(/V)| is not divisible by p, and so a simple c o u n t i n g argument suffices to find an A - i n v a r i a n t Sylow r-subgroup of N . ) Observe t h a t R is not n o r m a l in N since, otherwise, it w o u l d be characteristic, a n d hence n o r m a l i n G. But R 2 U, and so this w o u l d contradict the fact t h a t U is the unique m i n i m a l n o r m a l subgroup of G contained i n AT.
r

Now R U / U is a Sylow r-subgroup of the nilpotent group N / U , and thus R U / U < N / U and R U < N . A l s o , R U admits the action of A , and so R U is a solvable Frobenius kernel. If R U < N , then R U is nilpotent by the inductive hypothesis, and hence R is characteristic i n R U , and hence R < N , w h i c h is a contradiction. W e deduce t h a t R U = N . Since A R is a Frobenius group, it has a p a r t i t i o n II consisting of R a n d the \ R \ conjugates of A i n A R . T h e n | H | - 1 = \ R \ is a power of r , a n d so this number is coprime to \ U \ . B y L e m m a 6.8, therefore, some member X e II has n o n t r i v i a l fixed points i n the a b e l i a n group U. T h e action of every G-conjugate of A on U is Frobenius, however, a n d so X cannot be one of the conjugates of A . W e deduce t h a t X = R , and thus since U is abelian and R U = N , we have 1 < Cu(R) C Z{RU) = Z(N).

Since U is the unique m i n i m a l n o r m a l subgroup of G contained i n N and Z { N ) is n o n t r i v i a l and n o r m a l i n G, we conclude that U Q Z ( N ) , a n d thus R < U R = N . T h i s is a contradiction, and the proof is complete. N e x t , we state a result of T h o m p s o n t h a t w i l l enable us to prove the theorem of his thesis, w h i c h was t h a t a l l Frobenius kernels are nilpotent (and not just solvable Frobenius kernels, as i n T h e o r e m 6.22). A c t u a l l y , the m a i n result of T h o m p s o n ' s thesis was a c r i t e r i o n for a group to have a n o r m a l p-complement, where p is an o d d prime; the result about Frobenius kernels is a comparatively easy corollary. A b o u t five years after he completed his degree, T h o m p s o n found a much better version of his n o r m a l p-complement theorem w i t h a very m u c h easier proof, a n d t h a t is the result we present a n d

198

6. F r o b e n i u s

Actions

prove i n C h a p t e r 7. Here we give o n l y a p a r t i a l statement of this i m p r o v e d theorem. 6 . 2 3 . T h e o r e m ( T h o m p s o n ) . L e t P e S y \ { G ) , w h e r e G i s a finite g r o u p a n d p + 2, a n d assume t h a t N ( X ) has a n o r m a l p - c o m p l e m e n t f o r eve r y n o n i d e n t i t y c h a r a c t e r i s t i c s u b g r o u p X of P . T h e n G has a n o r m a l p complement.
p G

F i r s t , observe that this can be viewed as a strong form of Frobenius' normal p-complement theorem, w h i c h is our T h e o r e m 5.26. R e c a l l t h a t Frobenius' theorem tells us that to prove t h a t G has a n o r m a l p-complement, it suffices to show t h a t N ( X ) has a n o r m a l p-complement for every nonident i t y p-subgroup X of G. It is a t r i v i a l i t y (by Sylow theory) t h a t i n Frobenius' theorem, it is enough to consider o n l y p-subgroups X that are contained i n some fixed Sylow p-subgroup P of G. T h e force of T h e o r e m 6.23 is t h a t for p ^ 2, it suffices to restrict attention to nonidentity c h a r a c t e r i s t i c subgroups X of P ; it is not necessaty to consider a l l nonidentity subgroups of P . (The n o r m a l p-complement theorem of T h o m p s o n ' s thesis can also be viewed as a strong form of F r o b e n i u s ' theorem, v a l i d for p ^ 2. It is somewhat more c o m p l i c a t e d to state, however.)
G

We m e n t i o n t h a t the c o n d i t i o n p ^ 2 i n T h e o r e m 6.23 is essential. C o n sider, for example, G = 5 , the s y m m e t r i c group of order 24. A Sylow 2-subgroup P of G is isomorphic to the dihedral group D , and the nonidentity characteristic subgroups of P are P , the unique cyclic subgroup of order 4 i n P and Z ( P ) , of order 2. It is easy to check t h a t the normalizer in G of each of these subgroups is P itself, and so these normalizers a l l have n o r m a l 2-complements. T h e group G, however, does not have a n o r m a l 2-complement.
4 8

A s we shall see i n C h a p t e r 7, it is not a c t u a l l y necessary to consider a l l nonidentity characteristic subgroups X of P ; two of t h e m suffice. One of these is the center Z ( P ) , and the other is the T h o m p s o n subgroup J ( P ) , w h i c h we w i l l define. O f course, if P > 1, then Z ( P ) > 1, and as we shall see, also J ( P ) > 1. A s s u m i n g T h e o r e m 6.23, we now present the m a i n result of this section. 6 . 2 4 . T h e o r e m . L e t N be a F r o b e n i u s k e r n e l . Then N is nilpotent.

P r o o f . A s i n the proof of T h e o r e m 6.22, there exists a group A of p r i m e order p such t h a t A has a Frobenius action on N . P r o c e e d i n g by i n d u c t i o n on \ N \ , suppose first that there exists a n o r m a l A - i n v a r i a n t subgroup M of N w i t h 1 < M < N . T h e n the action of A on M is Frobenius, and by C o r o l l a r y 6.2, so too is the action of A o n N / M . B y the inductive hypothesis, therefore, b o t h M and N / M are nilpotent, and thus N is solvable. It follows

Problems 6 C

199

by T h e o r e m 6.22 that N is nilpotent, and we are done i n this case. W e can assume, therefore, that N contains no n o n t r i v i a l proper A - i n v a r i a n t n o r m a l subgroup. W e argue now that N is a n r - g r o u p for some p r i m e r , and thus N is nilpotent, as wanted. Otherwise, \ N \ has at least two p r i m e divisors, and so we can choose an o d d p r i m e r d i v i d i n g |7V|. L e t R be an A - i n v a r i a n t Sylow r-subgroup of N . (As i n the proof of T h e o r e m 6.22, the existence of a n A - i n v a r i a n t Sylow r-subgroup follows by a simple counting argument because | A | is prime.) L e t AT be a nonidentity characteristic subgroup of R. Since R is A invariant, it follows t h a t X is also A - i n v a r i a n t . F u r t h e r m o r e , X < N because N is not an r-group, and so by the result of the previous paragraph, X cannot be n o r m a l i n N . T h e n N ( X ) is a proper A - i n v a r i a n t subgroup of N , and so by the inductive hypothesis, N ( X ) is nilpotent, a n d hence it has a n o r m a l r-complement. ( E v e r y nilpotent group has a n o r m a l r-complement.) It follows by T h e o r e m 6.23 that N has a n o r m a l r-complement AT, and we have 1 < K < N since N is not an r-group, b u t it does have order divisible by r . A l s o , K is characteristic i n N , and so it is A - i n v a r i a n t . T h i s contradicts the result of the first paragraph, and so the proof is complete.
N N

Problems 6C 6 C . 1 . L e t a G A u t ( G ) , a n d suppose that a fixes only the identity i n G . (a) If a has prime order, show t h a t G is nilpotent. (b) Show by example that G need not be nilpotent if a has order 4. H i n t . T h e r e exists an example for (b) w i t h | G | = 75. 6 C . 2 . L e t A be elementary abelian of order p , a n d suppose that A acts on a nonnilpotent group N v i a automorphisms. A s s u m e t h a t C ( A ) = 1.
N 2

(a) Show t h a t N has has nilpotent subgroups K such that K i n K = 1 w h e n i ^ j .

5

> 1 for 1 < i < p + 1

(b) N o w fix a prime q, and let Q i be the unique Sylow g-subgroup of K where the AT; are the subgroups of (a). L e t X = ( Q i | 1 < i < p + l ) . Show t h a t X e S y l ( 7 V ) .
u g

H i n t . F o r (b), consider an A - i n v a r i a n t Sylow g-subgroup Q of N . Show first t h a t A C Q . If X < Q, show t h a t X C Y < Q, for some A - i n v a r i a n t subgroup Y t h a t is proper i n Q. Show t h a t the action of A on Q/Y is Frobenius.

Chapter 7

The Thompson Subgroup

7A In this chapter, we complete the proof of T h o m p s o n ' s theorem t h a t Frobenius kernels are nilpotent. T h e missing ingredient i n our p a r t i a l proof i n C h a p t e r 6 is to show that i f p ^ 2, then a sufficient c o n d i t i o n for a group G to have n o r m a l p-complement is t h a t for every n o n i d e n t i t y characteristic subgroup X of some Sylow p-subgroup P of G, the subgroup N ( X ) has a n o r m a l p-complement. (Since every subgroup of a group that has a normal p-complement must also have a n o r m a l p-complement, this is clearly a necessary c o n d i t i o n too, but, of course, t h a t fact is far less interesting.)
G

T h o m p s o n ' s first n o r m a l p-complement c r i t e r i o n appeared i n his 1959 Ph.D. thesis, w i t h a long and subtle proof. T h e n i n 1964, T h o m p s o n p u b lished a m u c h shorter (though still subtle) proof of a stronger theorem, a n d t h a t is the theorem a n d proof we present i n this chapter. ( T h o m p son's 1964 paper appeared i n the very first issue of the J o u r n a l of A l g e b r a , w h i c h was an auspicious beginning for t h a t periodical.) T h o m p s o n showed in his J o u r n a l of A l g e b r a paper that i f p + 2, then a group G has a n o r m a l p-complement p r o v i d e d t h a t N ( X ) has a n o r m a l p-complement for just two specific characteristic subgroups X of P , where P e S y l ( G ) . These two c r i t i c a l characteristic subgroups are the center Z ( P ) a n d the T h o m p s o n subgroup J ( P ) , b o t h of w h i c h are n o n t r i v i a l i f P is n o n t r i v i a l . (We know, of course, t h a t the center of a n o n t r i v i a l p-group is n o n t r i v i a l , a n d as we s h a l l see w h e n we present the definition, the T h o m p s o n subgroup of a n o n t r i v i a l p-group is also guaranteed to be nontrivial.) O f course, i f the normalizers
G p

201

202

7. T h e T h o m p s o n

Subgroup

of a l l n o n t r i v i a l characteristic subgroups of P have n o r m a l p-complements, then i n p a r t i c u l a r , N ( Z ( P ) ) a n d N ( J ( P ) ) have n o r m a l p-complements, and so T h o m p s o n ' s 1964 theorem yields our T h e o r e m 6.23, w h i c h was the key to the proof t h a t Frobenius kernels are nilpotent i n C h a p t e r 6. I n fact, T h o m p s o n ' s 1964 theorem is slightly stronger t h a n we have just stated since we c a n replace the normalizer of Z ( P ) w i t h the centralizer of Z ( P ) . (Since C ( Z ( P ) ) C N ( Z ( P ) ) , the requirement t h a t the centralizer should have a n o r m a l p-complement is weaker t h a n , a n d is i m p l i e d by, the corresponding assumption for the normalizer.)
G G G G

7.1. T h e o r e m ( T h o m p s o n ) . L e t P S y l ( G ) w h e r e G i s a finite g r o u p a n d p + 2, a n d assume that C (Z(P)) and N (J(P)) have n o r m a l p c o m p l e m e n t s . T h e n G has a n o r m a l p - c o m p l e m e n t .

p ; G G

A s we have explained, once we complete the proof of T h e o r e m 7.1, we w i l l have established t h a t a l l Frobenius kernels are nilpotent. T h e T h o m p s o n subgroup J ( P ) has other applications too, a n d i n particular, it provides a crucial ingredient i n the group-theoretic proof of B u r n s i d e ' s pV-theorem t h a t we present later i n this chapter. F o r that reason, we prove somewhat more about the T h o m p s o n subgroup t h a n the m i n i m u m needed for the proof of T h e o r e m 7.1. Unfortunately, the literature contains several slightly different definitions of "the" T h o m p s o n subgroup of a p-group P , a n d i n general, these definitions y i e l d different subgroups, a l l of w h i c h have been referred t o as J ( P ) . In particular, the version of J ( P ) t h a t we are about t o define is not always equal t o T h o m p s o n ' s J ( P ) . O u r definition is slightly easier t o use t h a n that in T h o m p s o n ' s paper, however, a n d we feel that this justifies the risk of confusion caused b y c o m p e t i n g definitions. G i v e n a p-group P , let (P) be the set of a l l of those elementary abelian subgroups of P t h a t have the m a x i m u m possible order. T h e n the T h o m p son s u b g r o u p J ( P ) of P is the subgroup generated b y a l l of the members of {P). F o r example, i f P is the d i h e d r a l group of order 8, then obviously, P does not contain an elementary abelian subgroup of order 8. B u t P does contain an elementary abelian subgroup of order 4, a n d so {P) is the set of a l l elementary abelian subgroups of order 4 i n P . There are two of these, and together they generate the whole group P , a n d thus J ( P ) = P i n this case. O f course, i f P is n o n t r i v i a l , then it must contain a n o n t r i v i a l element a r y abelian subgroup, for instance, one of order p , a n d thus the set (P) contains at least one n o n t r i v i a l subgroup. It follows that J ( P ) > 1, as we mentioned previously. In his 1964 paper, T h o m p s o n d i d not restrict attention t o elementary abelian subgroups. Instead, he defined J ( P ) to be the subgroup generated by

7A

203

all a b e l i a n subgroups of P of largest possible rank. ( T h e r a n k of an abelian p-group A is the integer r such that the unique largest elementary abelian subgroup i i i ( A ) has order p . E q u i v a l e n t l y , if the abelian p-group A is decomposed as a direct p r o d u c t of n o n t r i v i a l cyclic factors, then the r a n k of A is the number of such factors.) Since a n elementary abelian subgroup w i t h m a x i m u m possible order i n P is an abelian subgroup of m a x i m u m possible rank, we see t h a t our subgroup J ( P ) is always contained i n T h o m p s o n ' s , but it can be properly smaller. ( T h e containment is proper, for example, if P is abelian, but not elementary abelian.) Y e t another v a r i a t i o n o n the definition t h a t has appeared i n the literature is to consider the subgroup of P generated by a l l abelian subgroups of largest possible order.
r

A n i m m e d i a t e consequence of the definition is the following. 7.2. L e m m a . L e t 3 { P ) C Q C P , w h e r e P i s a p - g r o u p . J ( Q ) , a n d i n p a r t i c u l a r , J ( P ) i s c h a r a c t e r i s t i c i n Q. Then J(P) =

P r o o f . Since ((P)) = J ( P ) C Q, it follows that every m a x i m a l - o r d e r elementary abelian subgroup of P is contained i n Q, a n d because Q C P , these are m a x i m a l - o r d e r elementary abelian subgroups of Q. E v e r y m a x i m a l - o r d e r elementary abelian subgroup of Q, therefore, has the same order as the m e m bers of {P), a n d hence is a member of {P). It follows t h a t {Q) = (P), and thus 3 ( G ) = J ( P ) . Before we can begin our proof of T h e o r e m 7.1, we need a number of p r e l i m i n a r y results. T h e first of these is a fairly t e c h n i c a l l e m m a about the general linear group G L ( 2 , p ) , w h i c h , we recall, is the group of invertible 2 x 2 matrices over the field F of order p. 7 . 3 . L e m m a . L e t G = G L ( 2 , p ) , w h e r e p ^ 2 i s p r i m e , a n d l e t P C G be a p - s u b g r o u p . Suppose P C N ( L ) , f o r some s u b g r o u p L of G, w h e r e \ L \ i s n o t d i v i s i b l e by p , a n d assume f u r t h e r t h a t a S y l o w 2 - s u b g r o u p of L i s abelian. Then P C C (L).
G G

T h e condition that p ^ 2 i n L e m m a 7.3 is necessary since G L { 2 , 2 ) is isomorphic to the s y m m e t r i c group 5 , and so there is a p'-group L of order 3 n o r m a l i z e d but not centralized by a Sylow p-subgroup P of order 2. T h e somewhat u n n a t u r a l requirement t h a t L s h o u l d have an abelian Sylow 2subgroup also cannot be o m i t t e d , at least w h e n p = 3. If G = G L { 2 , 3), t h e n G has a subgroup L of order 8 that is n o r m a l i z e d b u t not centralized b y a Sylow 3-subgroup P , w h i c h has order 3. ( T h e p r o d u c t L P of order 24 is the special linear group S L ( 2 , 3), w h i c h is the set of matrices w i t h determinant 1 i n G.) T h i s is not a counterexample to L e m m a 7.3, however, since L is a nonabelian 2-group, isomorphic to the quaternion group Q . I n fact, it is only for p = 3 that such an example can occur, and so L e m m a 7.3 could be
3 8

204

7. T h e T h o m p s o n

Subgroup

strengthened by requiring the c o n d i t i o n on a Sylow 2-subgroup of L only if p = 3. W h e n we apply the l e m m a , however, we w i l l k n o w t h a t L is abelian, and so we have no need for a stronger result. W e digress briefly to discuss general linear groups i n general, and also special linear groups and certain other related groups. (We saw some of these groups i n C h a p t e r 1, and we study t h e m further i n C h a p t e r 8.) If F is a n a r b i t r a r y field and n is a positive integer, then G L ( n , F ) is the group of a l l invertible n x n matrices over F , and the special linear group S L { n , F ) is the subgroup of G L { n , F ) consisting of those matrices that have determinant equal to 1. Since the determinant m a p is a h o m o m o r p h i s m from G L ( n , F ) onto the m u l t i p l i c a t i v e group F of F , it follows t h a t S L ( n , F ) , w h i c h is the kernel of this h o m o m o r p h i s m , must be a n o r m a l subgroup. A l s o , we have G L { n , F ) / S L { n , F ) = F .
x x

N o w suppose t h a t F is finite, so that \ F \ = q, where q is a prime power. Since F is the unique field (up to isomorphism) of order q, it is no loss to write G L ( n , q ) i n place of G L ( n , F ) , and i n fact, this n o t a t i o n is fairly c o m m o n . Similarly, we w r i t e S L { n , q ) for the corresponding special linear group, w h i c h is a n o r m a l subgroup of index | F | = q - 1 i n G L ( n , q ) .
X

W e can compute \ G L { n , q ) \ by counting the number of ways we can construct a n n x n m a t r i x w i t h linearly independent rows, where each entry comes from the field F of order q. O f course, there are a t o t a l of q row vectors of length n over F . E a c h of these other t h a n the zero row is a possibility for the first row of our m a t r i x , and so there are exactly q - 1 possible first rows. T o keep the rows linearly independent, we must not allow the second row to be one of the q different scalar multiples of the (nonzero) first row, so after the first row is chosen, there are exactly q - q possibilities for the second row. After the first two (linearly independent) rows are selected, we must exclude a l l of their q different linear combinations from appearing as the t h i r d row, and this leaves q - q possibilities for the t h i r d row. C o n t i n u i n g like this, we see t h a t there are exactly q - q ~ possibilities for the k t h row, and we deduce t h a t
n n n 2 n 2 n k 1

n1

n
n

\ G L ( n , q ) \ = l\(q
i=0

- q ) = q

l[(qi

- I),

j=l

where N = 1 + 2 + + (n - 1) = n ( n - l ) / 2 . I n particular, | G L ( 2 , q ) \ = (q
2

- l)(q

- q ) = q ( q - \ ) { q + 1), a n d | 5 L ( 2 , ? ) | = q { q - l ) ( q + 1).

N o w consider the subgroup P C G L { n , q ) consisting of the matrices that have zeros below the diagonal, ones on the diagonal a n d a r b i t r a r y elements of F i n a l l above-diagonal positions. (Note that P really is a subgroup, a n d that P C S L ( n , q ) . ) Since row k of an n x n m a t r i x contains exactly k - 1 above-diagonal positions, we see that the t o t a l number of above-diagonal

7A

205

positions i n an n x n m a t r i x is 1 + 2 + + (n - 1) = N , where N is as before. T h e n \ P \ = q , a n d thus \ G L ( n , q ) \ = \ P \ m , where m is a product of numbers of the form g> - 1, a n d so m is relatively p r i m e to q. Since q is a power of some prime, say p , we see t h a t the subgroup P is a Sylow p-subgroup of G L { n , q ) . N o t e t h a t i f n > 1, we can easily find a second Sylow p-subgroup, different from P , by t a k i n g lower t r i a n g u l a r matrices i n place of the upper triangular subgroup P .
N

The scalar matrices i n G L ( n , q ) are the scalar multiples of the ident i t y m a t r i x , and so there are exactly q - 1 of these, and it is easy to see that these matrices form the center Z ( G L ( n , q ) ) . T h e factor group G L { n , q ) / Z ( G L { n , q ) ) is the projective general linear group, and is denoted P G L ( n , q ) . T h e scalar matrices that lie i n S L ( n , q ) are the matrices of the form a - 1 , where a = 1, and i n fact, these form the center Z ( 5 L ( 2 , g ) ) . W r i t i n g Z = Z { S L { n , q ) ) , we see t h a t \ Z \ is equal to the number d of elements a G F such that a = 1. Since F is cyclic of order q - 1, it follows t h a t d = ( n , q - 1), the greatest c o m m o n divisor. B y definition, the projective special linear group P S L ( n , q ) is the factor group S L { n , q ) / Z . For n > 2, this group is simple except w h e n n = 2 a n d q is 2 or 3. (In fact, P S L ( n , F ) is also simple when n > 2 and F is a n infinite field.)
n x n x

T a k i n g n = 2, we have \ P S L ( 2 , q ) \ = q ( q - l ) ( g + l ) / d , where d = 1 i f q is a power of 2 and d = 2 if q is o d d . I n p a r t i c u l a r , for p r i m e powers q = 4 , 5 , 7 , 8 , 9 , 1 1 we have \ P S L ( 2 , q ) \ = 6 0 , 6 0 , 1 6 8 , 5 0 4 , 3 6 0 , 6 6 0 , respectively, a n d these account for a l l of the n o n a b e l i a n simple groups of order at most 1000. (We have not o m i t t e d the alternating groups A and As since P S L ( 2 , 4 ) = A s = P S L { 2 , 5 ) a n d P S L { 2 , 9 ) = A . ) W e m e n t i o n also t h a t P S L { 2 , 7 ) = P S L { 2 > , 2 ) a n d A = P 5 L ( 4 , 2 ) . I n fact, these are the o n l y isomorphisms among alternating groups a n d the simple groups of type P S L ( n , q ) . (The simple groups P 5 L ( 3 , 4 ) a n d A have equal orders, but they are not isomorphic.)
5 6 8 8

F i n a l l y , we m e n t i o n one way that general linear groups arise i n abstract group theory. C o n s i d e r a n elementary a b e l i a n p-group E of order p , and view E as a n additive group. T h e n E can be identified w i t h a vector space of dimension n over the field of order p , and under this identification, we see that A u t ( F ) = G L { n , p ) . T h i s observation (in the case n = 2) is crucial to the proof of T h e o r e m 7.1, and that is w h y L e m m a 7.3 is relevant.
n

In order to prove L e m m a 7.3, we need the following observation about the groups S L { 2 , q ) , where q is o d d .

7.4.

L e m m a . If q i s o d d ,

t h e n t h e n e g a t i v e of t h e i d e n t i t y m a t r i x i s t h e

unique involution i n SL{2,q).

206

7. T h e T h o m p s o n

Subgroup

T h i s can be proved by a c o m p u t a t i o n a l argument along the following lines. If t is an i n v o l u t i o n i n S L { 2 , q ) , then the fact that t = 1 yields four nonlinear equations that must be satisfied by the four entries of the m a t r i x t, a n d the fact t h a t det(t) = 1 yields one more equation. It is not h a r d to show t h a t if the characteristic is different from 2, these five equations have just two solutions, corresponding to t = I and t = - I , where I is the 2 x 2 identity m a t r i x . W e prefer the following more conceptual proof, however.
2

P r o o f o f L e m m a 7 . 4 . L e t t S L ( 2 , q ) be a n i n v o l u t i o n other t h a n - / . W r i t e f { X ) to denote the p o l y n o m i a l X - 1, a n d observe t h a t f { t ) = 0. Since neither the p o l y n o m i a l X - 1 nor the p o l y n o m i a l X + 1 yields 0 w h e n t is substituted for X , it follows that f ( X ) = X - l is the m i n i m a l p o l y n o m i a l for t. T h e C a y l e y - H a m i l t o n theorem asserts that the m i n i m a l p o l y n o m i a l of an a r b i t r a r y square m a t r i x divides the characteristic p o l y n o m i a l , and i n our situation, where t is a 2 x 2 m a t r i x , we know that the characteristic p o l y n o m i a l of t has degree 2, and hence f ( X ) is the characteristic p o l y n o m i a l of t. T h e n t has two distinct eigenvalues, 1 a n d - 1 , and so the p r o d u c t of the eigenvalues of M s - 1 . B u t the p r o d u c t of the eigenvalues (counting m u l tiplicities) for a n a r b i t r a r y square m a t r i x is the determinant of the m a t r i x , and it follows that det(t) = - 1 . B u t t 6 S L { 2 , q ) , and so det(t) = 1, and this c o n t r a d i c t i o n completes the proof.
2 2

It follows by T h e o r e m 6.11 that a Sylow 2-subgroup of S L { 2 , q ) (for o d d q ) is either cyclic or generalized quaternion. A n elementary argument t h a t shows that the cyclic case can never occur depends on the fact t h a t i n a finite field F of o d d order q, it is always possible to find elements a and b such t h a t a + b = - 1 . A s s u m i n g that a a n d b satisfy this c o n d i t i o n , consider the two matrices:
2 2

0 1

-1 0

and

a b b a

These clearly have determinant 1, and so they lie i n S L { 2 , q ) . It is easy to check t h a t x = - I = y and that y = -y = y ~ \ and it follows t h a t the subgroup ( x , y ) of S L ( 2 , q ) is isomorphic to the quaternion group Q . A Sylow 2-subgroup of S L { 2 , q ) cannot be cyclic, therefore, a n d so it must be generalized quaternion. W e m e n t i o n also that since \ S L ( 2 , 3)| = 24, a Sylow 2-subgroup of S L ( 2 , 3) is isomorphic to Q .
2 2 x 8 8

To see w h y the equation a + b = - 1 must have a solution i n F , observe t h a t the number of elements of F that have the form 1 + a is exactly ( q + l ) / 2 . ( T h i s is so because the m a p a i-> 1 + a is two-to-one for nonzero elements of F , a n d so the image of this m a p contains (q - l ) / 2 elements different from 1, for a t o t a l of (q + l ) / 2 elements.) Similarly, there are e x a c t l y ( q + l ) / 2 elements of F that have the form - b , and since the
2 2 2

7A

207

sum of the cardinalities of these two sets exceeds g, the sets must overlap. It follows that 1 + a = - b
2 2

for some choice of a , b F , as wanted.

P r o o f o f L e m m a 7 . 3 . W o r k i n g by i n d u c t i o n on \ L \ , we can assume t h a t P centralizes every proper subgroup of L t h a t it stabilizes. If we choose a p r i m e q d i v i d i n g | L : C ( P ) | , we can find a P - i n v a r i a n t S y l o w g-subgroup Q of L ,
L

a n d since Q % C L ( P ) , we must have Q = L , a n d thus L is a g-group. A l s o , [L,P] is P - i n v a r i a n t , a n d so i f [ L , P ] < L , we have [ L , P , P ] = 1, a n d since [ L , P , P ] = [ L , P ] by L e m m a 4.29, we conclude t h a t [L, P ] = 1, as required. We m a y assume, therefore, t h a t [ L , P] = L , a n d i n p a r t i c u l a r , L C G' C SL{2,p), abelian. If q = 2, t h e n by assumption, the 2-group L is abelian, a n d we k n o w by L e m m a 7.4 that L contains a unique i n v o l u t i o n . It follows that L is a cyclic 2-group, a n d so A u t ( L ) is also a 2-group, a n d the p-group P cannot act n o n t r i v i a l l y on L . W e can assume, therefore, t h a t q is an o d d p r i m e , and since \ L \ is a power of q that divides \ S L ( 2 , p ) \ = p { p - l ) ( p + 1), it follows t h a t \ L \ divides one of p - 1 or p + 1. ( T h i s , of conrse, is because the p r i m e q cannot divide b o t h p - 1 a n d p + 1.) I n p a r t i c u l a r , we have \ L \ < p + 1. N o w i f P acts n o n t r i v i a l l y on L , there must be at least one P - o r b i t i n L of size at least p , a n d together w i t h the identity, this yields \ L \ > p + 1. W e conclude that \ L \ = p + 1, a n d so \ L \ is even. B u t \ L \ is a power of the o d d p r i m e q, a n d this c o n t r a d i c t i o n completes the proof. where the second containment follows because GL(2,p)/SL(2,p) is isomorphic to the m u l t i p l i c a t i v e group of the field of order p , a n d so it is

N e x t , we prove what we call the " n o r m a l - P t h e o r e m " . 7.5. the T h e o r e m . L e t P G Syl (G), w h e r e G i s a
p

finite

group,

and

assume

following

conditions.

(1) G i s p - s o l v a b l e . (2) p + 2. (3) A Sylow (4) G acts (5) Then The \ V : C

v

2-subgroup faithfully ( P ) \ < p .

of G i s a b e l i a n . p-group V.

o n some

P < G . reader m a y wonder how T h e o r e m 7.5, w h i c h applies only to p -

solvable groups, can possibly be relevant to the p r o o f of T h o m p s o n ' s theorem (7.1), where there certainly is no p-solvability a s s u m p t i o n . T h e answer is that the proof of T h e o r e m 7.1 proceeds v i a a subtle i n d u c t i o n that w i l l guarantee p-solvability when we need it. ( T h o m p s o n ' s i n d u c t i v e argument

208

7. The Thompson

Subgroup

w i l l also guarantee that the Sylow 2-subgroup is a b e l i a n w h e n that fact is needed.)

P r o o f o f T h e o r e m 7 . 5 . A s s u m e t h a t G is a counterexample of m i n i m u m possible order. T h e n G does not have a n o r m a l Sylow p-subgroup, a n d we can two G, choose Q S y l ( G ) w i t h Q + P . T h e subgroup ( Q , P ) of G has at least
p

Sylow p-subgroups, a n d it clearly satisfies a l l five hypotheses of the we have G = Now Q = P

9

theorem. It too is a counterexample, therefore, a n d so by the m i n i m a l i t y of (P,Q). for some element g G, a n d thus C { Q )
V V V

{C (P))
V

also has index at most p i n V. L e t U = C ( P ) \V : U \ < \ V : and C {P)\\V

V

n C { Q ) , a n d observe that

:C (Q)\
V

< p , a n d U < V. A l s o , since b o t h P a n d

Q act t r i v i a l l y o n U, it follows that the action of G = ( P , Q ) o n U is t r i v i a l , so [ U , G ] = 1. In p a r t i c u l a r , U is G - i n v a r i a n t , a n d we have a n a t u r a l V/U. so that [V, K ] C U, a n d C [ U , K ] = 1. Since V is a p-group, it follows that K must (If Q is a Sylow g-subgroup of AT, where q ^ p, t h e n Let K be the kernel of the a c t i o n of G o n V/U, action of G o n thus [ V , K , K ]

also be a p-group.

[V, Q] = [V, Q, Q] = 1, a n d thus Q = 1 since the a c t i o n of G on V is faithful, by assumption.)_Since K C O ( G ) , it follows that K C P _ a n d A ^ C Q , a n d

p

thus the group G = G / K acts faithfully o n V/U, index \ V : C { P ) \

V

has distinct Sylow p-subgroups P a n d Q . A l s o , G a n d we see that P centralizes C ( P ) / U ,

V

w h i c h has V/U,

< p i n V/U.

It follows t h a t the group G satisfies a l l five

conditions w i t h respect to its Sylow p-subgroup P a n d its a c t i o n o n therefore, we must have K = I , a n d so G acts faithfully o n V/U. thus replace V by V / U a n d assume t h a t \V\ < p .
2

but it does not have a n o r m a l Sylow p-subgroup. B y the m i n i m a l i t y of G , W e can

Since G acts faithfully o n V, it is isomorphically embedded i n A u t ( V ) . If V is cyclic, then A u t ( V ) is abelian, a n d so G is a b e l i a n a n d P < G, w h i c h is not the case. W e conclude that V is noncyclic, a n d so it must be elementary abelian of order p , a n d thus A u t ( V ) = G L { 2 , p ) , a n d we can view G as a subgroup of G L { 2 , p ) .
p 2

I n p a r t i c u l a r , \P\ < p , a n d since P is not n o r m a l , N o w let L = C y ( G ) , so t h a t P normalizes

it follows t h a t O ( G ) = 1. the p'-group L . trivially on L .

G

Since we are assuming that a Sylow 2-subgroup of G is B y the H a l l - H i g m a n L e m m a 1.2.3 (our T h e o r e m 3.21) we

abelian, the same is true for L , a n d thus by L e m m a 7.3, we see that P acts have P C C ( L ) C L , a n d thus since P is a p-group a n d L is a p'-group, we have P = 1. T h i s is a c o n t r a d i c t i o n since P is not n o r m a l .

Problems 7 A

209

Problems 7A 7A.1. Show t h a t a nilpotent m a x i m a l subgroup of a simple group must be a 2-group. N o t e . T h e simple group PSL{2,17) has a m a x i m a l subgroup of order 16, so n o n t r i v i a l 2-groups actually c a n occur as m a x i m a l subgroups of simple groups. 7A.2. L e t S = 5 L ( 2 , 3 ) a n d w r i t e Z = ( - / ) , so t h a t Z is the unique subgroup of order 2 i n S. Show t h a t the group S/Z of order 12 has four S y l o w 3-subgroups, and deduce t h a t it has a unique Sylow 2-subgroup. C o n c l u d e from this t h a t S has a n o r m a l subgroup of order 8. H i n t . Use the fact mentioned i n the text t h a t i f n > 2 a n d q is a power of the p r i m e p , then G L { n , q ) has at least two S y l o w p-subgroups. N o further m a t r i x c o m p u t a t i o n s are required for this p r o b l e m . 7A.3. L e t G = G L ( n , q ) , where q is a power of the p r i m e p , a n d let P be the Sylow p-subgroup of G consisting of the upper t r i a n g u l a r matrices w i t h ones on the diagonal. L e t D be the group of a l l diagonal matrices i n G. (a) Show t h a t D C N ( P ) .
G

(b) Show that D P is the group of a l l upper t r i a n g u l a r matrices i n G w i t h a r b i t r a r y nonzero entries on the diagonal. (c) V i e w G as acting by right m u l t i p l i c a t i o n on the space of n-dimensional row vectors over the field F of order q. Identify a l l P - i n v a r i a n t subspaces of V, and observe t h a t there is exactly one such subspace of dimension k for each integer k w i t h 1 < k < n . (d) Show t h a t a l l of the P - i n v a r i a n t subspaces of V are TV-invariant, where N = N ( P ) , and use (c) to deduce t h a t TV = D P .
G

7A.4.

If q is a power of the p r i m e p , show t h a t S L { 2 , q ) a n d P S L ( 2 , q ) each

have exactly q + 1 Sylow p-subgroups. 7A.5. L e t P be a p-group, a n d suppose t h a t U < P is elementary abelian. Show t h a t U normalizes some group Ee(P). H i n t . Choose E e {P) such that | P n E \ is as large as possible. If U does not normalize E , w r i t e P = E , where u U and E ^ F . L e t H = ( E , F ) and Z = E n P . Show that Z ( H n U ) G {P).
u

7A.6. Suppose t h a t Q is a generalized quaternion group a n d that | A u t ( Q ) | is divisible by an o d d p r i m e p . Show t h a t p = 3 a n d \ Q \ = 8, a n d deduce t h a t the hypothesis t h a t a Sylow 2-subgroup is abelian i n L e m m a 7.3 a n d T h e o r e m 7.5 is unnecessary if p > 3.

210

7. T h e T h o m p s o n

Subgroup

7B In this section, we prove the following " n o r m a l - J theorem". Burnside's p V - t h e o r e m . 7.6. T h e o r e m . L e t P e S y l ( G ) , w h e r e G i s a the f o l l o w i n g conditions.
p

T h i s w i l l be

used for the proof of T h e o r e m 7.1, and also for the group-theoretic proof of

finite

group, and

assume

(1) G i s p - s b l v a b l e . (2) p ^ 2 . (3) A S y l o w 2 - s u b g r o u p (4) C y ( G ) = l . (5) P = C ( Z ( P ) ) .

G

of G i s a b e l i a n .

T h e n J(P) < G. Since we are assuming t h a t G is p-solvable a n d t h a t C y ( G ) - 1, it follows (assuming that G is n o n t r i v i a l ) that O ( G ) > 1. T h e point of the theorem, however, is not s i m p l y t h a t G has a n o n t r i v i a l n o r m a l p-subgroup; it is t h a t a p a r t i c u l a r p-subgroup, namely J ( P ) , is n o r m a l i n G . N o t e that it is not u n t i l Step 7 of the following proof that we use the a c t u a l definition of the set {P).
p

P r o o f of T h e o r e m 7.6. Suppose t h a t the theorem is false, a n d let G be a counterexample of m i n i m u m possible order, p a r t i c u l a r , G is n o n t r i v i a l , so O p ( G ) > 1, and we w r i t e U = O ( G ) and G = G / U . A l s o , let I = C y (G), where L D U . (This, of course, unambiguously defines the subgroup L < G.) W e proceed i n a number of steps, the first of w h i c h consists of three closely related statements, a l l of w h i c h are consequences of the H a l l - H i g m a n L e m m a 1.2.3.
p

Step 1. (a) Z ( P ) C U. (b) If P C P C G , then O / ( P ) = 1.

p

(c) C Q ( L ) C L . P r o o f . Since G is p-solvable a n d C y ( G ) = 1, L e m m a 1.2.3 (our Theorem 3.21) tells us t h a t U = O ( G ) D C ( P ) . B u t U C P , a n d so Z(P) C C ( U ) a n d (a) follows. A l s o , since U = O ( G ) , we see that O (G) = 1, a n d L e m m a 1.2.3 yields (c). F o r (b), let U C H C G and write M = ( V ( i f ) . T h e n M a n d U are n o r m a l i n H a n d M n U = 1 since U is a p-group and M is a p'-group. T h u s M C C { U ) C U, so M = M n P = 1, p r o v i n g (b).
p G G p p G

7B

211

Step 2. T h e r e exists A G (P)

such that A % U.

P r o o f . Otherwise, a l l members of (P) are contained i n P , and thus J ( P ) C P . It follows by L e m m a 7.2 that J ( P ) = J ( P ) is characteristic i n P , and since U < G , we have J ( P ) < G . T h i s is a c o n t r a d i c t i o n since we are assuming that G is a counterexample. Step 3. L e t U A C H < G , and suppose t h a t P n P G S y l ( P ) .
p

Then A

centralizes P / T l L . Proof. First, P p is stilt of H is satisfies Observe that P satisfies the first four hypotheses of the theorem. is p -solvable since it is a subgroup of the p -solvable group G . A l s o , different from 2, so the second c o n d i t i o n holds. A Sylow 2-subgroup contained i n a Sylow 2-subgroup of G , w h i c h is abelian, and so P c o n d i t i o n (3), a n d finally, O , { H ) = 1 by Step 1(b).
p p

Now let S = H n P G S y l ( P ) . Since Z ( P ) C i / C S C P b y Step 1(a), we have Z ( P ) C Z ( 5 ) , and thus C ( Z ( S ) ) C C ( Z ( P ) ) = P. Thus C j y ( Z ( 5 ) ) is a p -subgroup of H c o n t a i n i n g the S y l o w p -subgroup S, and so C ( Z ( S ) ) = S and H satisfies the fifth hypothesis too.
H G H

Since H < G , the theorem holds for H , a n d hence J ( 5 ) < P . A l s o , since A G ( P ) and A C 5 C P , it follows t h a t A G (S), and so A C J ( 5 ) . T h e n [PnL,A] c [PnL,J(5)] c ( H n L ) n J ( S ) = LnJ(S) c p,

where the second containment holds because b o t h P n L and J ( 5 ) are n o r m a l in P , and the final containment holds because U is the unique Sylow p subgroup of L , and thus it contains every p-subgroup of P W e now have 1 = [ P n L , A] = [ P n L , A], as wanted. Step 4. G = L A and P = P A . Proof. Write P = L A , and observe t h a t P A is a p -subgroup of P and

t h a t |P : P A | = \ L ( U A ) : U A \ = \ L : L n P A | ,
p

w h i c h divides the p '-number

| L : P | . It follows that P A G S y l ( P ) , and thus U A = P n P . If H < G , then since L C P , Step 3 yields A C C ^ ( L ) C L , where the second containment holds by Step 1(c). Since A is a p -group a n d L is a p '-group, we have A = 1, and so A C P . T h i s contradicts the choice of A , however, and we conclude that P = G , as wanted. F i n a l l y , we have U A = H n P Step 5. | A | = p. P r o o f . F i r s t , A is n o n t r i v i a l since A % U. A l s o , A is elementary abelian, and so it suffices to show that it_is cyclic. N o w A j i c t s c o p r i m e l y on L , and this action is faithful since C Q ( L ) C L and L n A = 1. B y L e m m a 6.20, = G n P = P .

212

7. T h e T h o m p s o n

Subgroup

therefore, it suffices to show that A acts t r i v i a l l y on every A - i n v a r i a n t proper subgroup of L . Suppose t h a t M is ^ - i n v a r i a n t , where M < L , and where we assume (as we may) t h a t M D U. T h e n A C N ( M ) , and so M A is a group, and M A D U A P . A l s o , since A is a p-group, the p'-part of \ M A \ is equal to the p'-part of | M | , and this is s t r i c t l y less t h a n the p'-part of \ L \ since \ L : M | is a p'-number exceeding 1. It follows that M A < G, and we can a p p l y Step 3 to deduce that A centralizes M A n L D M , p r o v i n g Step 5.
G

Now let V = { z G Z ( P ) | = 1}, so that V is a n elementary abelian n o r m a l subgroup of G. T h e n G acts by conjugation on F , j m d since V C Z ( P ) , we see t h a t J7 acts trivially. T h i s yields an a c t i o n of G = G / U on V. Step 6. T h e action of G on V is faithful. P r o o f . L e t K - C ( V ) , so that K is the kernel of the a c t i o n of G on V. W e argue that AT is a p-group by considering a S y l o w g-subgroup Q of AT, where g is any p r i m e different from p. T h e n Q acts coprimely o n the abelian group Z(E7), and Q fixes a l l elements of order p i n Z ( U ) since these elements a l l lie i n V and Q C AT - C ( V ) . It follows by C o r o l l a r y 4.35 t h a t Q acts t r i v i a l l y on Z ( U ) . B u t Z ( P ) C U, and so Z ( P ) C Z([7), and thus Q C C ( Z ( P ) ) = P . W e conclude that Q = 1, and thus AT is a p-group, as claimed. B u t AT < G , so AT C O ( G ) = 17, and thus AT = 1, as wanted.
G G G p

Step 7. | V : V n A \ < p.

7C

213

P r o o f . W r i t e D = U T \ A and E = V C \ A as i n the d i a g r a m , a n d observe t h a t \ V : E \ = \ V : V n D \ = \ V D : D \ . N o w D is a n elementary abelian subgroup of P , and V is a central elementary abelian subgroup of P , and thus V D is elementary abelian. Since A 6 (P), it follows that < | A | , and thus |VD : P | < \ A : P | = = p. W e now have \ V \ E \ = \ V D : D \ < p , as wanted. Step 8. W e have a contradiction. We want to apply the " n o r m a l - P theorem" ( T h e o r e m 7.5) to the faithful action of G on V, and so we need to check t h a t \ V : C ( P ) \ < p . B u t P = U A , so P = A , and of course, A centralizes V n A since A is abelian. T h u s \ V : C ( P ) \ < \ V : V n A \ < p , as wanted, and we conclude t h a t P < G. T h e n P < G and A C P C O ( G ) = U, w h i c h is not the case.
V V p

7C In this section, we prove T h e o r e m 7.1, thereby c o m p l e t i n g the proof t h a t Frobenius kernels are nilpotent. W e b e g i n w i t h a n extension of L e m m a 2.17. 7.7. L e m m a . W r i t e G = G/N, w h e r e N i s a n o r m a l p ' - s u b g r o u p of a finite

g r o u p G. If P i s a p - s u b g r o u p o f G , w e h a v e (a) N Q ( P ) = N (b) C Q ( P ) = P r o o f . Statement

G

( P ) and

CoJP). (a) is essentially L e m m a 2.17. A l s o , since overbar is a

G

h o m o m o r p h i s m , it is clear that C

( P ) C C ( P ) , a n d so to establish (b), it
G

suffices to prove the reverse c o n t a i n m e n t . _ B y (a), overbar defines a surjective h o m o m o r p h i s m from N G _ ( P ) to N ^ P ) , a n d so by the correspondence theorem, the subgroup C ^ ( P | C r % ( P ) is the image of some subgroup X

N (P).
G

I n other words, X = C ^ ( P ) , a n d we have 1 = [ P , X ] = [P, X ] ,

G

and thus [ P , X ] C N . B u t X C N ( P ) , a n d so we also have [P, X ] C_ P . T h u s [ P , X ] C N n P = 1, and X C C ( P ) .

G

W e conclude that C ^ P ) =

A C C ( P ) , and the proof is complete.

G

W e are now ready to prove T h e o r e m 7.1, w h i c h , we recall, asserts that G has a n o r m a l p-complement if p ^ 2 and b o t h N ( J ( P ) ) a n d C ( Z ( P ) ) have n o r m a l p-complements, where P is a S y l o w p-subgroup of G. (Of course, i f these conditions h o l d for P , they w i l l also h o l d for every Sylow p-subgroup of G.) W e repeatedly use the fact that the p r o p e r t y of having a n o r m a l p-complement is inherited by subgroups and h o m o m o r p h i c images.
G G

214

7. T h e T h o m p s o n

Subgroup

Observe the subtle way that the subgroup U is chosen i n the following proof. T h i s is one of the keys to this very clever argument. ( T h i s trick appears b o t h i n T h o m p s o n ' s thesis and i n his 1964 paper.) P r o o f of T h e o r e m 7 . 1 . Suppose that G is a counterexample of m i n i m a l order. T h e n G fails to have a n o r m a l p-complement, and so by F r o b e n i u s ' theorem ( T h e o r e m 5.26), there must be some nonidentity p-subgroup U of G such t h a t N ( P ) fails to have a n o r m a l p-complement. A m o n g a l l such "bad" subgroups, choose P such that | N ( P ) | is as large as possible. (In other words, U is chosen so that a Sylow p-subgroup of N ( P ) has order as large as possible.) Subject to t h a t condition, choose U to have order as large as possible. W e proceed i n a number of steps.
G G p G

Step 1. P =

O (G).
p p G

P r o o f . L e t S 6 S y l ( 7 V ) , where N = N ( P ) , and suppose t h a t N < G. T h e n N is not a counterexample to the theorem, and since N fails to have a n o r m a l p-complement, at least one of N ( 3 ( S ) ) or C ( Z ( S ) ) must fail to have a n o r m a l p-complement. T h e n one of N ( J ( S ) ) or C ( Z ( 5 ) ) fails to have a n o r m a l p-complement, and so by hypothesis, S cannot be a full Sylow p-subgroup of G. T h e n S is properly contained i n a Sylow p-subgroup, and hence we can choose a p-subgroup T such that S < T and S < T. (We are, of course, using the fact t h a t "normalizers grow" i n p-groups.)
N N G G

Now N ( A ) fails to have a n o r m a l p-complement, where X is one of J (S) or Z ( 5 ) . A l s o U > 1, so S > 1, and thus X > 1, a n d therefore X is a member of the set of b a d subgroups from w h i c h we selected P . A l s o , X is characteristic i n S, and so X < T and T C N ( A ) . W e thus have | N ( X ) | > | T | > \S\ = | N ( P ) | , and since this contradicts the choice of P , we conclude t h a t N = G, and so U < G and U C O ( G ) . N o w O ( G ) is a member of the set of nonidentity p-subgroups whose normalizers fail to have n o r m a l p-complements, and since | N ( O ( G ) ) | = \ P \ = | N ( [ / ) | , the choice of U guarantees t h a t \ U \ > | O ( G ) | , and thus U = O ( G ) , as required.
G G G P G p p p G p p G p p p

Step 2. G / U has a n o r m a l p-complement, Proof. Write G = G / U ,

and so G is p-solvable. T h e n G is not a

and observe t h a t J G | < \ G [ .

p

counterexample to the theorem, and so since P 6 S y l ( G ) , it suffices to show t h a t each of N ( J ( P ) ) and C ( Z ( P ) ) has a n o r m a l p-complement. T h e r e
G G

is n o t h i n g to prove i f p does not d i v i d e X P

5

and so we can assume that P is

n o n t r i v i a l , and thus J ( P ) a n d Z ( P ) are n o n t r i v i a l . L e t U C X C P , where is either J (J ) or Z ( P ) , and observe that X > U. A l s o , X < P , and so C N
G

( X ) and | N ( X ) |
G

= |P| = |N (P)| .
G P

B u t | A | > | P | , and so by

the choice of P , we see that X cannot be one of the b a d subgroups whose

7C

215

normalizers fail to have n o r m a l p-complements. Since X contains the kernel U of o v e r b a d e have N ^ X ) = N

G

( X ) by the correspondence

theorem,

a n d thus N ^ X ) has a n o r m a l p-complement. I n p a r t i c u l a r , b o t h r % ( J ( P ) ) a n d N g ( Z ( P ) ) have n o r m a l p-complements, a n d therefore, G has a n o r m a l p-complement, as required. Step 3. C y ( G ) = 1. P r o o f . L e t K = C y ( G ) , a n d w r i t e G G/K. P. N o w P is a S y l o w p-subgroup

of G , a n d since K is a p'-group, overbar defines an i s o m o r p h i s m from P onto It follows that J ( P ) = J ( P ) and Z ( P ) = Z ( P ) . T h e n

NG(J(P)) = Ng(JCP)) = N ( J ( P ) )
G

and

CG(Z(P)) = CG(Z(P)) =

C (Z(P)),
G

where the second e q u a l i t y m each of these_equations follows by L e m m a 7.7. W e conclude that I % ( J ( P ) ) a n d C ^ { Z { P ) ) _ T h i s is not the case, however, since K required. Step 4. P is a m a x i m a l subgroup of G. P r o o f . Suppose that P C H < G. Then N ( J ( P ) ) and C ( Z ( P ) )
H H

have n o r m a l p-complements, a n d G fails to have a

a n d so i f \ G \ < | G | , it w o u l d follow that G has a n o r m a l p-complement. = O ,{G)

p

n o r m a l p-complement. T h e n | G | is not smaller t h a n \ G \ , a n d so K = 1, as

have

n o r m a l p-complements, a n d since H is not a counterexample, it follows that H has a n o r m a l p-complement K .

P

T h e n K < H and O ( G ) < H ,

P G P P

a n d since K n O ( G ) = 1, we have K C C ( O ( G ) ) C O ( G ) .

(We are

using L e m m a 1.2.3, w h i c h applies because G is p-solvable by Step 2 a n d C y (G) = 1 by Step 3.) T h e n K = I a n d P is a p-group, a n d hence P = P . Step 5. C ( Z ( P ) ) = P .
G

P r o o f . C e r t a i n l y , C ( Z ( P ) ) 5 P . B u t C ( Z ( P ) ) < G since G fails to have

G G

a n o r m a l p-complement, a n d thus C ( Z ( P ) ) = P b y Step 4.

G

Step 6. T h e n o r m a l p-complement of G / P is abelian. P r o o f . F i r s t , recall t h a t G / P has a n o r m a l p-complement L / U Suppose t h a t P C X C L , where X is n o r m a l i z e d by P . by Step 2.

T h e n P X is a

group, a n d hence by Step 4, either P X = P or P X = G . If P X = P , t h e n | X | is a power of p, a n d so X = P . If, o n the other h a n d , P X = G , t h e n | G : X | is a power of p, a n d hence X = L . It follows t h a t no n o n i d e n t i t y proper subgroup of I = L / P is P - i n v a r i a n t .

216

7. T h e T h o m p s o n

Subgroup

We c a n certainly assume that L is n o n t r i v i a l , and so it has a n o n t r i v i a l P - i n v a r i a n t Sylow g-subgroup for some prime q. It follows from the result of the previous paragraph that L is a g-group, and since it is n o n t r i v i a l , we have ( I ) ' < L . O f course, ( I ) ' is P - i n v a r i a n t , and thus ( I ) ' = 1, and it follows that L is abelian. Step 7. W e have a c o n t r a d i c t i o n . P r o o f . W e argue that G satisfies a l l five hypotheses of the n o r m a l - J theorem ( T h e o r e m 7.6). Certainly, p ^ 2, and we know that G is p-solvable, that O > ( G ) = 1, and that P = C ( Z ( P ) ) . A l l that remains, therefore, is to check t h a t a Sylow 2-subgroup of G is abelian.
p G

WriteJL7 = G / U and let Q G S y l ( G ) , so that Q is a p'-subgroup of G , and thus Q is contained i n the abelian n o r m a l p-complement of G. (We are using Step 6 here.) A l s o , Q = Q because P is the kernel of the canonical h o m o m o r p h i s m G - > G and Q n P = 1, and thus Q is abelian, as wanted. B y T h e o r e m 7.6, therefore, J ( P ) < G . B u t then G = N ( J ( P ) ) has a n o r m a l p-complement, and this is our final contradiction.
2 G

Problems 7C 7C.1. ( T h o m p s o n ) L e t P G S y l ( G ) , where G is finite and p ^ 2, and suppose t h a t N ( A ) / C ( A ) is a p-group for every characteristic subgroup X of P . Show t h a t G has a n o r m a l p-complement.
p G G

H i n t . W o r k b y i n d u c t i o n on | G | and use the inductive hypothesis to show t h a t if X is characteristic i n P and A > 1, then N ( X ) has a n o r m a l p-complement.
G

7D In this section, we prove B u r n s i d e ' s theorem. 7 . 8 . T h e o r e m . L e t G be a p r i m e s . Then G is solvable. finite g r o u p of o r d e r p q ,

a b

where p and q are

Burnside's proof of this theorem was one of the first major applications of character theory, early i n the 20 t h century. (Characters were invented by Frobenius, who used t h e m to prove his theorem on the existence of Frobenius kernels, but B u r n s i d e was one of the early developers of the theory.) B u r n s i d e ' s proof appears i n the second (1911) e d i t i o n of his group theory text, of w h i c h a very substantial fraction is devoted to character a n d l i n ear representation theory. It is interesting that the first (1897) edition of B u r n s i d e ' s b o o k includes no character theory at a l l , apparently because he

7D

217

felt that it was not good for m u c h . B u r n s i d e wrote i n his preface to the first e d i t i o n that "it w o u l d be difficult to find a result t h a t could be most directly obtained b y the consideration of [linear representations]." H e began the preface to the second edition w i t h the assertion t h a t "Very considerable advances ... have been made since the appearance of the first e d i t i o n of this book." (Was he t h i n k i n g p r i m a r i l y of his o w n pV-theorem?) In fact, B u r n s i d e ' s proof is elegant a n d not very difficult; it lies relatively near the surface of character theory, a n d it appears fairly early i n most m o d e r n texts on the subject. Nevertheless, it has always seemed somewhat strange t h a t one h a d to use facts about the field of c o m p l e x numbers i n order to prove this theorem about finite groups. B y the late 1960s, considerable advances (to use B u r n s i d e ' s words) h a d been made i n pure finite group theory. I n p a r t i c u l a r , the T h o m p s o n subgroup, a n d T h o m p s o n ' s techniques of "local analysis" h a d been developed, and these h a d proved to be very powerful tools, as we have seen i n this chapter. T h o m p s o n ' s P h . D . student D a v i d G o l d s c h m i d t wondered whether or not these tools could be used to give a non-character proof of B u r n s i d e ' s theorem. ( T h e real point here was to demonstrate the power of T h o m p son's techniques; no one claimed that there was any deficiency i n B u r n s i d e ' s elegant proof.) G o l d s c h m i d t almost succeeded: he was able to prove the p Y - t h e o r e m p r o v i d e d t h a t b o t h p a n d q are different from 2. A few years later, a n d a p p r o x i m a t e l y simultaneously, H . Bender a n d H . M a t s u y a m a found ways to handle the r e m a i n i n g case. M a t s u y a m a ' s argument was easier, b u t B e n der's approach, w h e n supplemented b y M a t s u y a m a ' s idea, yielded a n overall simplification of G o l d s c h m i d t ' s argument, a n d t h a t is essentially the proof we present here. ( B u t Bender used G l a u b e r m a n ' s Z ( J ) - t h e o r e m , w h i c h we have decided not to include i n this book. Instead, we appeal to the n o r m a l - J theorem.) P r o o f o f T h e o r e m 7 . 8 . L e t G be a counterexample of smallest possible order. Since the order of every proper subgroup of G also has at most two p r i m e divisors, every such subgroup must be solvable, a n d also i f 1 < N < G, then G / N is solvable. I n particular, i f AT is a n o n t r i v i a l proper n o r m a l subgroup of G, then b o t h N a n d G / N are solvable, a n d hence G is solvable, w h i c h contradicts our assumption t h a t G is a counterexample. It follows t h a t G is simple. W e focus now o n the m a x i m a l subgroups of G. If AT is any n o n t r i v i a l n o r m a l subgroup of a m a x i m a l subgroup M , t h e n of course, M C N ( A T ) , and since 1 < AT < G a n d G is simple, we k n o w t h a t N ( A T ) < G. It follows by the m a x i m a l i t y of M t h a t M = N ( K ) . ( T h i s observation w i l l be used repeatedly i n what follows.) A l s o , i f M is m a x i m a l , then M is solvable
G G G

218

7. T h e T h o m p s o n

Subgroup

and n o n t r i v i a l , and so it has a nonidentity n o r m a l subgroup of prime-power order. T h u s either O ( M ) > 1 or O ( M ) > 1, and one of our intermediate goals is to show that it is not possible for b o t h O ( M ) and O ( M ) to be n o n t r i v i a l . W e proceed i n a number of steps.
p q p q

Step 1. Suppose that K C G is nilpotent, and assume t h a t M = N ( A T ) is a m a x i m a l subgroup of G . If b o t h p and q divide \ K \ , then M is the unique m a x i m a l subgroup of G containing K .
G

P r o o f . A s s u m i n g that the assertion is false, let AT be m a x i m a l among counterexample subgroups, and let AT C X , where A is a m a x i m a l subgroup of G different from M . W r i t e K = K x K , where K a n d K are respectively the (nontrivial) Sylow p-subgroup and Sylow g-subgroup of AT. Since K is characteristic i n AT, we have K < M , and it follows t h a t M = N ( A T ) .
p q p q p p G p

Now M i l X = N ( A T ) is a p-local subgroup of X . A l s o , K < M n X , and so K C C y (AT n X ) . Since X is solvable and X n M is a p-local subgroup of X , we can a p p l y T h e o r e m 4.33 to conclude that C y ( M n X ) C C y ( X ) . T h i s yields K C C y ( X ) = O { X ) , and so w r i t i n g L = O ( X ) , we have K C L . Similarly, K C L , where L = O ( X ) , and thus b o t h L and L are n o n t r i v i a l . W r i t e L = L L , and note that K C L . A l s o , L is nilpotent, | L | is divisible by b o t h p and g and X = N ( L ) is a m a x i m a l subgroup of G. If AT < L , it follows by the choice of K t h a t X is the unique m a x i m a l subgroup of G that contains L . B u t AT C L , and thus
x p q q q q q q q q p p p p p q p q G P p

A
p

C C ( L ) C C (AT ) C N ( A ) = M ,
G P G P G p

and similarly, L C M . T h e n L C M , and since X is the unique m a x i m a l subgroup c o n t a i n i n g L , we have M = X , w h i c h is a c o n t r a d i c t i o n . In the r e m a i n i n g case, K = L , and thus M = N ( A T ) = N ( L ) = X , and again we have a c o n t r a d i c t i o n .
G G

Step 2. L e t P e S y l ( G ) , and suppose t h a t P normalizes some n o n t r i v i a l subgroup V of G. T h e n (V, Q ) = G for every Sylow g-subgroup Q of G. I n particular, V cannot be a g-group.
p

P r o o f . L e t Q G S y l , ( G ) , and write H = ( V , Q ) . N o w | P Q | = | P | | Q | = | G | since P n Q = 1, and thus P Q = G . L e t g be an a r b i t r a r y element of G and write g = xy, w i t h x G P and y G Q. Since P normalizes V , we have V = V - V C P , and thus P contains a l l G-conjugates of V . It follows t h a t H contains the nonidentity n o r m a l subgroup V generated by the conjugates of V, and since G is simple, V = G, and so P = G . T o prove the final assertion, observe t h a t if V is a g-group, we can choose Q to contain V, and then (V, Q ) = Q < G , c o n t r a d i c t i n g what we have proved.
9 xy y G G

Step 3. L e t M be m a x i m a l i n G . T h e n either O ( M ) = 1 or O ( M ) = 1.
p q

7D

219

P r o o f . A s s u m e t h a t the assertion is false, a n d w r i t e Z = Z ( O ( M ) ) and Z = Z ( 0 ( M ) ) , so that Z and Z are b o t h n o n t r i v i a l and n o r m a l i n M . A l s o , Z and Z are abelian, and they centralize each other since they intersect t r i v i a l l y , and thus the group Z = Z Z is abelian. Furthermore, Z < M , and so M = N ( Z ) , and it follows by Step 1 t h a t M is the unique m a x i m a l subgroup of G that contains Z . N o w i f 1 ^ z G Z , then Z C C (z) < G , and so C (z) is contained i n some m a x i m a l subgroup of G that contains Z , and we conclude that C { z ) C M .
p p g ? p q p q p q G G G G

Let 5 G S y l ( M ) . Since S normalizes the n o n t r i v i a l g-subgroup Z , it follows by Step 2 t h a t S cannot be a full S y l o w p-subgroup of G . T h e n 5 < P for some Sylow p-subgroup P of G , a n d hence N ( S ) is a p-subgroup of G s t r i c t l y larger t h a n S. T h e n N ( S ) M , and we can choose a n element g G G - M such that 5 = S.
p q P P

Now w r i t e A = ( Z y , and observe that since Z C O ( M ) C 5 , we have ^4 = ( Z ) 9 C S = S C M , and thus A acts on the n o r m a l subgroup Z of M . Suppose either t h a t the a c t i o n of A o n Z is not faithful or that A is not cyclic. I n b o t h cases, we argue t h a t
p p p 9 p q q

= (C

Z q

{a) \ l ^ a e A ) .

T h i s is clear if the a c t i o n is not faithful since i n t h a t case, there exists a nonidentity element a G A such that C ( a ) = Z . If A is not cyclic, then since it is abelian, T h e o r e m 6.20 applies to y i e l d the assertion.
Z q q

We argued previously t h a t C (z) C M for a l l n o n i d e n t i t y elements z G Z . T h e n C { z ) C Af for a l l n o n i d e n t i t y elements z G Z , and i n particular, since A = ( Z ) C 2 J , w e have C ( a ) C M i f l ^ o e A If the action of A o n Z is not faithful or if A is not cyclic, therefore, we deduce that Z C M f . A l s o Z C S = S C M , a n d thus Z = Z Z C M . B u t M is the unique m a x i m a l subgroup c o n t a i n i n g Z , a n d therefore M = M and we have 5 G N ( M ) = M , w h i c h is not the case.
G 9 9 9 G p G 9 q p p ? 9 G

We deduce t h a t A is cyclic and t h a t it acts faithfully on Z . A l s o , since A = ( Z y , we see that Z is cyclic. B y similar reasoning, w i t h the roles of p and q interchanged, we deduce t h a t Z is cyclic. B u t A is a n o n t r i v i a l p-group that acts faithfully on Z , and so p divides | A u t ( Z , ) | , and since Z is cyclic, we see that p divides q - 1, a n d so p < q. E x p l o i t i n g the s y m m e t r y between p and q once again, we have q < p , and this, of course, is a c o n t r a d i c t i o n , w h i c h completes the proof of Step 3.
q p p q q q

If M is a n a r b i t r a r y m a x i m a l subgroup of G , we have observed t h a t at least one of O ( M ) or O ( M ) is n o n t r i v i a l , and now by Step 3, we k n o w t h a t exactly one of these subgroups is n o n t r i v i a l . I n other words, G has two flavors of m a x i m a l subgroups: those for w h i c h O ( M ) > 1 and those for w h i c h O ( M ) > 1, and we refer to these as p - t y p e m a x i m a l s a n d g - t y p e
p q p q

220

7. T h e T h o m p s o n

Subgroup

m a x i m a l s , respectively. W e stress that every m a x i m a l subgroup of G has exactly one of these types. Before we proceed w i t h the next step, we define a p - c e n t r a l element of G to be a n o n i d e n t i t y element i n the center of some Sylow p-subgroup of G, and similarly, a g - c e n t r a l element is a nonidentity element of the center of some Sylow g-subgroup of G. Step 4. Suppose t h a t y G N ( V ) , where y is g-central and V is a p-subgroup. T h e n V contains no p-central elements.
G

P r o o f . G i v e n a p-subgroup U C G, write U* to denote the subgroup generated by the p-central elements i n U. Since the G-conjugates of a p-central element are p-central, it follows t h a t N ( J 7 ) permutes the p-central elements of U, and thus N ( J 7 ) normalizes U*. In p a r t i c u l a r , since y normalizes V, it is also true that y normalizes V*, and our goal is to show t h a t V* = 1.
G G

Now V* is a p-subgroup of G t h a t is n o r m a l i z e d by y and is generated by p-central elements, and we let W be a p-subgroup of G, m a x i m a l subject to the conditions that it is n o r m a l i z e d by y and generated by p-central elements. If V* > 1, then W > 1, and we work to o b t a i n a c o n t r a d i c t i o n . W r i t e TV = N ( W ) , and let S G Syl (TV) so that S Z } W . Since y is g-central, we can choose Q G S y l ( G ) such that y G Z ( Q ) . T h e n (y) is n o r m a l i z e d by Q, and since y G TV, we have (y, S) C TV < G , where the strict inequality holds because 1 < W < TV.
G p ?

It follows by Step 2 ( w i t h the roles of p and g reversed) t h a t S cannot be a full Sylow p-subgroup of G . Since S G Syl (TV), we can reason as before that N ( 5 ) g and thus there exists an element g e G - TV such that
p G

= S. I n particular, W

^ ^ 7 , but V P C S

= S CN .
9

Since VP is generated by p-central elements and W at least one p-central generator x of W such t h a t x
p 9

+ W, there must be 0 W.
5

Because x is pafe

central, we can choose P e S y l ( G ) such that x G Z ( P ) . N o w G = P Q , and so we can write g = a b , w i t h a G P and b G Q . T h e n x and x
5

= x

= x

G W

C TV, and thus x

GW
b

n TV. Since W
b

n TV is a p-subgroup n TV) is a p-subgroup

b

t h a t normalizes the p-group W, it follows that W(W of G . A l s o , W = W*, and so ( W ( W

6 b

n TV))*
6

D (W,x) >

Now V P is n o r m a l i z e d by y = y . Since y G TV, it follows that W D TV is n o r m a l i z e d by y, and therefore ( ^ ( V F n N ) ) * is a p-subgroup that is n o r m a l i z e d b y y. T h i s subgroup is generated by p-central elements and it s t r i c t l y contains W, c o n t r a d i c t i n g the choice of W. Step 5. E v e r y p subgroup of G is centralized by a p-central element. A l s o , p-type m a x i m a l subgroups of G contain no g-central elements.

7D

221

P r o o f . G i v e n a p-subgroup V C G, choose P G S y l ( G ) w i t h P D F . T h e n Z ( P ) C C ( V ) , and so C ( V ) contains a p-central element.

p G G

Now suppose t h a t M is a p-type m a x i m a l subgroup, and let V = O ( M ) , so that V > 1 and 0 ( M ) = 1 by Step 3. T h e n V 5 C ( V ) = C ( V ) b y the H a l l - H i g m a n L e m m a 1.2.3, and thus V contains a p-central element. B y Step 4, no g-central element lies i n N ( V ) = M , as required.
p 9 M G G

Step 6. A g-central element of G cannot normalize a n o n t r i v i a l p-subgroup. P r o o f . L e t V be a n o n t r i v i a l p-subgroup of G, a n d let M be a m a x i m a l subgroup of G containing N ( V ) . It follows by Step 5 t h a t M contains a p-central element, a n d thus by Step 5 again, w i t h p a n d q interchanged, M cannot have g-type, and thus it has p-type. B y Step 5 yet again, M cannot contain a g-central element, and i n p a r t i c u l a r , N ( V ) contains no g-central element.
G G

Step 7. p + 2 and g ^ 2. P r o o f . Suppose g = 2 a n d choose a n i n v o l u t i o n t i n the center of a Sylow g-subgroup. B y T h e o r e m 2.13, there exists a n element x G G of order p such t h a t x = X ' , and thus t G N ( ( x ) ) . Since i is g-central, however, this contradicts Step 6. W e deduce t h a t g ^ 2, and, similarly, p ^ 2.
l 1 G

Step 8. L e t M be a p-type m a x i m a l subgroup of G, and let S G S y l ( M ) .

p

T h e n J ( 5 ) < M a n d 5 is a full Sylow p-subgroup of G. P r o o f . W e w i s h to a p p l y the n o r m a l - J theorem ( T h e o r e m 7.6) to the group M , a n d so we proceed to check the five hypotheses. F i r s t , M is solvable, and so it is certainly p-solvable. B y Step 7, we k n o w that p ^ 2 and t h a t a Sylow 2-subgroup of M is t r i v i a l , a n d hence is abelian, and thus the second and t h i r d hypotheses h o l d . Since M has p-type, we k n o w t h a t C y ( M ) = O ( M ) = 1, and the fourth c o n d i t i o n holds.
q

The r e m a i n i n g c o n d i t i o n is t h a t C ( Z ( S ) ) = S, and to establish this, it suffices to show that C ( Z { S ) ) is a p-group. Otherwise, C ( Z ( S ) ) contains some subgroup Y of order q, and thus Z ( S ) normalizes Y. T h i s w i l l contradict Step 6 ( w i t h p and q interchanged) if we can show that Z ( S ) contains a p-central element.
M M M

Let S C P G S y l ( G ) , and observe t h a t S = M n P . Since M is a p-type m a x i m a l subgroup, we can write M = N ( V ) for some p-subgroup V, and we have V C S C P . T h e n Z ( P ) C N ( V ) f~]P = M C \ P = S, a n d it follows t h a t Z ( P ) C Z { S ) . T h u s Z ( S ) contains p-central elements, a n d so C ( Z ( 5 ) ) = S, as wanted. W e see now t h a t M satisfies the hypotheses of the n o r m a l - J theorem, so we have J ( 5 ) < J M .
p G G M

222

7. T h e T h o m p s o n

Subgroup

F i n a l l y , i f S is not a full Sylow p-subgroup of M , we can write S < T , where T is a p-subgroup of G properly containing S. T h e n T N ( J ( 5 ) ) = M , and this is a c o n t r a d i c t i o n since T > S and 5 is a full Sylow p-subgroup of M .
G

Step 9. W e have a c o n t r a d i c t i o n . P r o o f . O f course, the p-part and the g-part of \ G \ are unequal, and so we can assume that \ G \ > \ G \ . If S,T G S y l ( G ) , therefore, we have
P q p

|G |
P

> \ G \ \ G \ = \ G \ > \ST\ =

p q

|o n T\ and so | S T l T | > 1.
p p

= .fill, , \s n T |

Since O ( G ) = 1 and J ( P ) > 1 for P G S y l ( G ) , the T h o m p s o n subgroups of the Sylow p-subgroups of G cannot a l l be equal, so we can choose S,T G Sylp(G) w i t h J (5) ^ J ( T ) . D o this so that D = S n T is as large as possible, and observe that D < S since S ^ T. B y the c o m p u t a t i o n of the first paragraph, D > 1, a n d hence N ( D ) < G , a n d we can choose a m a x i m a l subgroup M D N ( D ) . N o w N ( D ) contains a p-central element by Step 5, a n d thus by Step 5 w i t h p and q reversed, M cannot be of g-type. T h e n M is of p-type, and so by Step 8, the Sylow p-subgroups of M are full Sylow p-subgroups of G .
G G G

Choose U G S y l ( M ) w i t h U D M n 5 , a n d observe t h a t
p

f/nS

= [ / n M n S = M n S D N ( P ) > D.
5

Since P G S y l p ( G ) , it follows by the m a x i m a l i t y of \ D \ that J ( 5 ) = J (17), and similarly, we can choose V G S y l ( M ) such that 3 ( V ) = J ( T ) . N o w J(P) and J ( V ) are conjugate i n M since the Sylow subgroups P and F are conjugate. B y Step 8, however, J ( P ) < M , and we deduce that J ( 5 ) = J ( P ) = J ( V ) = J ( T ) . T h i s contradicts the choice of S and T, and the proof is complete.
p

Chapter 8

Permutation Groups

8A In this chapter, we r e t u r n to the s t u d y of group actions on abstract sets. Let G act on a nonempty finite set ft, a n d recall t h a t such an a c t i o n defines a n a t u r a l h o m o m o r p h i s m (the p e r m u t a t i o n representation associated w i t h the action) from G into the s y m m e t r i c group S y m ( f t ) . W e say t h a t G is a p e r m u t a t i o n g r o u p on ft i f the a c t i o n is faithful, or equivalently, if the p e r m u t a t i o n representation is a n injective h o m o m o r p h i s m . If G is a p e r m u t a t i o n group on ft, it is c o m m o n to identify G w i t h its isomorphic copy i n S y m ( f t ) , so one can view p e r m u t a t i o n groups on ft s i m p l y as subgroups of S y m ( f t ) . O f course, if we intend to use a p e r m u t a t i o n representation of G to make statements about the group G itself, it is essential t h a t the corresponding action s h o u l d be faithful, but for m a n y purposes, a n d , i n particular, for the basic definitions of the subject, faithfulness is irrelevant. For that reason, and despite the title of this chapter, we w i l l generally not assume t h a t our actions are faithful except w h e n such an a s s u m p t i o n is really necessary. W e w i l l always assume i n this chapter, however, t h a t the sets being acted on are finite, although m u c h of the theory works more generally. R e c a l l that if G acts on ft, then ft is p a r t i t i o n e d into G - o r b i t s , and that the action is transitive if there is just one orbit: ft itself. E q u i v a l e n t l y , the action of G on ft is transitive i f for every choice of points a , 5 G ft, there exists a n element g G G such that a - g = 3. T h e following describes the (necessarily nonempty) set of a l l elements of G t h a t carry a to 0 i n a transitive action. (It is essentially a restatement of T h e o r e m 1.4, w h i c h underlies the fundamental c o u n t i n g principle.)

223

224

8. P e r m u t a t i o n

Groups

8.1. L e m m a . Suppose a G ft, a n d l e t H = G P G ft, l e t TT{P) G, cosets ofH.

a

t h a t a g r o u p G acts

t r a n s i t i v e l y o n ft. F i x a p o i n t TT{P) i s a r i g h t coset of H i n

be the s t a b i l i z e r of a i n G. F o r a n a r b i t r a r y p o i n t \ x G G } of a l l r i g h t

= { x G G I a-x = P ) . Then

a n d IT i s a b i j e c t i o n f r o m ft o n t o the set A = { H x

P r o o f . G i v e n 0 G ft, choose g G G such t h a t a-g = p , a n d observe that since the elements of H stabilize a , every element of the coset H g carries a to
P ,

a n d hence H g C i r ( P ) . Conversely, i f x G i r ( 0 ) , t h e n a-x = p , a n d

l

so a - { x g ~ )

= a . T h e n , xg~
TT

G H , a n d hence x G H g , a n d it follows that

= TT{^),

ir(p)

= H g G A. The map is injective since if i r { P ) t h e n a n element of this set

carries a to P a n d also to 7, a n d thus p = 7. F i n a l l y , TT is surjective since for each coset H y G A , we have H y = i r ( a - y ) . U

Of course, if G is an a r b i t r a r y finite group a n d H C G is any subgroup, t h e n G acts t r a n s i t i v e l y o n the right cosets of H i n G. I n the s i t u a t i o n of L e m m a 8.1, therefore, G acts on A , a n d it is interesting to compare this action w i t h the given a c t i o n of G on ft. I n fact, the bijection TT : ft -> A of L e m m a 8.1 is a p e r m u t a t i o n i s o m o r p h i s m , w h i c h means that i r ( P - g ) = n { P ) - g for a l l p G ft a n d g G G. (The "dot" on the right of this equation, of course, denotes the right m u l t i p l i c a t i o n a c t i o n of G on right cosets of H . ) T o check that the right cosets of H on each side of the equation x G G is an a r b i t r a r y element that carries a to p . It should now be apparent that the study of the transitive actions of a group is equivalent to the s t u d y of the actions of the group o n the sets of right cosets of its various subgroups. Furthermore, we k n o w t h a t the kernel of the a c t i o n of G o n the set of right cosets of a subgroup H is exactly c o r e ( A ) , a n d thus the s t u d y of transitive p e r m u t a t i o n groups (that is,
G 7

ir(0-g) = i r ( P ) - g
where

are the same, it suffices to observe that each of t h e m contains xg,

faithful transitive actions) is equivalent to the study of subgroups h a v i n g t r i v i a l core. I n other words, m u c h of permutation-group theory can be subsumed w i t h i n "pure" group theory. Nevertheless, the p e r m u t a t i o n point of view provides useful insights a n d suggests interesting theorems, a n d that is why we study it. A l s o , one should remember that historically, p e r m u t a t i o n groups (as subgroups of s y m m e t r i c groups) appeared before C a y l e y gave the m o d e r n a x i o m a t i c definition of a n abstract group. F o r m a n y years, the p e r m u t a t i o n group point of view was the only one available i n group theory. R e c a l l that i f G acts on ft a n d g G G a n d a G ft, t h e n ( G ) 3 = G . .
a a g

In

a transitive action, therefore, the point stabilizers are a l l conjugate, and, i n p a r t i c u l a r , if some point stabilizer is t r i v i a l , t h e n a l l of t h e m are t r i v i a l .

SA

225

If the point stabilizers i n a transitive action of G on ft are t r i v i a l , it follows by L e m m a 8.1 that for each choice of a,P G ft, there is exactly one element of G t h a t carries a to (3. I n this situation, where for a l l a and 13, there is a unique element g G G such t h a t a-g = P, we say that G is regular or s h a r p l y transitive on ft. (Conversely, of course, if G is regular on ft, t h e n there is a unique group element that carries a to a , and so the point stabilizer G is the t r i v i a l subgroup.) N o t e t h a t regular actions are a u t o m a t i c a l l y faithful.
a

B y the fundamental counting principle, we see that if G is transitive on ft, then | G | is a m u l t i p l e of |ft|. (The integer |ft| is the degree of the action.) In particular, \ G \ > |ft|, and G is regular if and o n l y if equality holds. W e m e n t i o n t h a t a not-necessarily-transitive a c t i o n i n w h i c h a l l point stabilizers are t r i v i a l is said to be semiregular. The n o t i o n of t r a n s i t i v i t y can also be refined i n a different direction. Suppose t h a t G acts on a set ft c o n t a i n i n g at least k points, where k is some positive integer, and consider the set > (ft) of ordered fe-tuples (ai, a , , a ) of distinct points on G ft. (The c o n d i t i o n |ft| > k is needed so t h a t O (ft) w i l l be nonempty.) T h e a c t i o n of G on ft defines a n a t u r a l componentwise a c t i o n of G on O { Q ) v i a the formula
fc 2 k fc k

(ai, a

, a

) - g

= {{a^-g,

( a ) - g , ( a
2

) - g ) .

(It is a t r i v i a l i t y that this is a genuine group action.) W e say t h a t G is A ;-transitive on ft if the associated componentwise action of G on O (ft) is transitive. Equivalently, and continuing to assume that |ft| > k, an a c t i o n of G on ft is fc-transitive if given any two ordered fe-tuples { a , a , . . . , a ) a n d { ( 3 i , ( 3 , . . . , ( 3 ) of distinct points of ft, there exists a n element g G G such t h a t ( a i ) - g = P i for 1 < i < k. (Note that a l t h o u g h we require t h a t the are distinct and t h a t the p i are distinct, there is no requirement t h a t the sets and { # } should be disjoint.)
fc x 2 k 2 k a i

Observe that i f G is fc-transitive on ft, t h e n G is a u t o m a t i c a l l y m transitive for a l l integers m w i t h 1 < m < k. I n particular, a fc-transitive a c t i o n is necessarily 1-transitive, w h i c h s i m p l y means t h a t it is transitive, a n d thus i n general, /c-transitivity is stronger t h a n transitivity. A n a c t i o n t h a t is fc-transitive w i t h k > 1 is said to be m u l t i p l y transitive, a n d if k = 2, the action is d o u b l y transitive. I n particular, m u l t i p l y transitive actions are always d o u b l y transitive. If |ft| = n , it is easy to see t h a t | O ( f t ) | = n ( n - 1) (n - k + 1), where there are exactly k factors i n the product. If G is A;-transitive o n ft, therefore, then because it is transitive on O ( f t ) , the fundamental counting principle guarantees that | G | must be a m u l t i p l e of | O ( f t ) | . I n particular, if G is d o u b l y transitive of degree n , then n ( n - 1) must divide | G | .
fc fc fc

226

8. P e r m u t a t i o n

Groups

W h a t are some examples of m u l t i p l y transitive group actions? F i r s t , it is easy to see t h a t the full s y m m e t r i c group S is n-transitive on ft = { 1 , 2 , . . . , n } . I n fact, given two ordered n-tuples (OH) and (#) of distinct points of ft, there is obviously a unique element g G S that carries o n to P i for a l l i. I n general, if a n a c t i o n is A;-transitive, and if for each pair of fc-tuples of distinct points there is a unique element of G c a r r y i n g one to the other, we say that the action is s h a r p l y fc-transitive. I n other words, G is sharply fc-transitive o n ft precisely when it is sharply transitive ( i . e . , regular) o n the set Ojt(ft). I n particular, if G is sharply fc-transitive on ft, where |ft| = n , then | G | = \O ()\ = n ( n l ) ( n - 2 ) ( n k + 1).
n n k

The alternating group A is ( n - 2)-transitive on ft = {1, 2 , . . . , n } . T o see this, observe t h a t the stabilizer of the (n - 2)-tuple t = (1, 2 , . . . , n - 2) in the s y m m e t r i c group S contains only the identity a n d the transposition exchanging the points n - 1 a n d n . T h e stabilizer of t i n A , therefore, is t r i v i a l , a n d so by the fundamental counting principle, the A - o r b i t of t has c a r d i n a l i t y equal to \ A \ . B u t \ A \ = n ! / 2 = | 0 _ ( f t ) | , a n d it follows that A is transitive o n 0 _ ( f t ) as wanted. I n fact, since a point stabilizer i n the a c t i o n of A o n 0 _ ( f t ) is t r i v i a l , A is sharply transitive on this set of (n - 2)-tuples, and so A is sharply (n - 2)-transitive on ft.
n n n n n n n 2 n n 2 n n 2 n n

We m e n t i o n that A a n d S are the o n l y subgroups of S that are (n - 2)-transitive on ft = { 1 , . . . , n } . One way to see this is to observe that if G S is (n - 2)-transitive, then |<D _ (ft)| = n ! / 2 must divide \ G \ , and thus G has index at most 2 i n S . It follows that G contains every element of order 3 i n S , and, i n particular, it contains a l l 3-cycles. It is not h a r d to see, however, that the 3-cycles i n S generate A , and thus A G. In other words, G is A or S , as claimed.
n n n n n 2 n n n n n n n

We have seen t h a t S and A are respectively sharply n-transitive and (n - 2)-transitive of degree n . I n fact, unless k is tiny, there are only a few other examples (sharp or not) of fc-transitive p e r m u t a t i o n groups. One of these is the M a t h i e u group M , w h i c h is one of the 26 sporadic simple groups. It is sharply 5-transitive on 12-points, and so
n n 1 2

\M

1 2

\ = 1 2 - 1 M 0 - 9 - 8 = 95,040 = 2 - 3 - 5 - l l .

The stabilizer of a point i n M is the sporadic simple group M , w h i c h is sharply 4-transitive on the r e m a i n i n g 11 points. (In general, we w i l l see t h a t i f k > 1, then a point stabilizer i n a fc-transitive action o n n points necessarily acts ( k - I n t r a n s i t i v e l y on the r e m a i n i n g n - 1 points, and it is easy to check that sharpness is inherited i n this situation.) O f course M has index 12 i n M , a n d so
1 2 n n 1 2

|Mn|

= 1 M 0 - 9 - 8 = 7,920 = 2 - 3 - 5 - l l .

8A

227

In fact, M n is the smallest of the sporadic simple groups. A point stabilizer in M is sharply 3-transitive on the r e m a i n i n g 10 points, but it is not simple. T h i s group, w h i c h is sometimes called M , has order 7920/11 = 720. It has a n o r m a l subgroup of index 2, isomorphic to the a l t e r n a t i n g group A , but M i o is n o t isomorphic to S . I n fact, there are three nonisomorphic groups of order 720 that contain copies of A ; these are S , M and P G L { 2 , 9 ) . ( R e c a l l that P G L { n , q ) is the projective general linear group; it is the factor group G L { n , q ) / Z , where *Z = Z ( G L ( n , q ) ) is the group of scalar matrices in G L { n , q ) . ) E a c h of these three groups of order 720 happens to be a p e r m u t a t i o n group of degree 10. B o t h M a n d P G L ( 2 , 9 ) are 3-transitive in their 10-point actions, but S is only 2-transitive a n d not 3-transitive. (One way to distinguish these three groups is v i a their Sylow 2-subgroups: a Sylow 2-subgroup of S is a direct p r o d u c t of D w i t h a cyclic group of order 2; a Sylow 2-subgroup of P G L { 2 , 9) is D a n d a Sylow 2-subgroup of M i o is the semidihedral group S D . )
n w 6 6 6 6 w w 6 6 8 m m

The M a t h i e u group M (which is another of the 26 sporadic simple groups) is 5-transitive o n 24-points. B u t it is not sharply 5-transitive; the stabilizer i n M of a 5-tuple of distinct points has order 48. It follows t h a t
2 4 2 4

|M |
2 4

= 4 8 | C | = 48-24-23-22-21-20 = 244, 823,040 = 2 -3 -5-7-11.23 .

5 2 4 2 3

10

The stabilizer of a point i n M is the M a t h i e u group M , w h i c h is also one of the sporadic simple groups; it is 4-transitive on the r e m a i n i n g 23 points. Clearly, M has index 24 i n M , and so | M | = 10,200,960. T h e stabilizer of a point i n M is the simple M a t h i e u group M , w h i c h is 3transitive on the remaining 22 points and has order 443, 520. T h e stabilizer of a point i n M is simple, but it is not one of the sporadic simple groups; it is isomorphic to P 5 L ( 3 , 4 ) . ( A s a consequence of L e m m a 8.29, we shall see t h a t F 5 L ( 3 , 4 ) is a 2-transitive p e r m u t a t i o n group of degree 21.)
2 3 2 4 2 3 2 3 2 2 2 2

For k > 4, we have named four fe-transitive p e r m u t a t i o n groups other t h a n S and A i n their n a t u r a l actions. These four groups are M and M , w h i c h are 5-transitive, and M and M , w h i c h are 4-transitive. In fact, there are no other examples. T h e only k n o w n proofs of this, however, require an appeal to the classification of simple groups. ( B u t there is an elementary argument due to C . J o r d a n that shows t h a t a s h a r p l y fe-transitive group w i t h k > 4 must either be a full s y m m e t r i c or a l t e r n a t i n g group, or else k is 4 or 5 a n d the degree is 11 or 12, respectively.) W e shall see, however, that there are infinitely m a n y 3-transitive p e r m u t a t i o n groups other t h a n s y m m e t r i c and alternating groups, a n d i n fact there are infinitely m a n y s h a r p l y 3-transitive groups. (See P r o b l e m 8A.17.)
n n 2 4 1 2 2 3 u

8.2. L e m m a . Suppose t h a t a g r o u p G acts t r a n s i t i v e l y o n ft. L e t 7 6 f t , a n d fix a n i n t e g e r k < |ft|. T h e n G i s k - t r a n s i t i v e o n ft if a n d o n l y if t h e p o i n t s t a b i l i z e r G i s ( k - \ ) - t r a n s i t i v e o n ft - {7}.

7

228

8. P e r m u t a t i o n

Groups

P r o o f . F i r s t , assume that G is fc-transitive o n ft. L e t a = {a ,a ,...,a . )

1 2 k 1

and

b = ( J 3 p , ,
U 2

Pk-i)

be (fc-l)-tuples of distinct points of ft-{7}, and let A and B be the fc-tuples constructed by appending 7 to a and b , respectively. T h e entries i n a and in b a r e distinct a n d lie i n ft - {7}, a n d hence the entries i n each of A and B are distinct. B u t G is fe-transitive on ft, so there exists an element g G G that carries A to B . T h e n g carries a to b a n d 7 to 7, or, i n other words, a - g = b and G G . It follows that G is ( k on ft - {7}, as wanted.

Conversely, suppose that G

2

is ( k

- Intransitive - Intransitive
7

on ft - {7}, and let

2

A = ( i , a , ,fe)

and

B = {Pi, p , . . . ,

P)
k

be ordered fc-tuples of distinct points of ft. Since the action of G on ft is transitive, we c a n choose elements x , y G G such t h a t ( a ) - s = 7 and = 7. T h e n A - x a n d are fc-tuples of distinct points of ft, and the last entry of each of t h e m is 7.
fc

Let a and b be the ( f c - l ) - t u p l e s obtained from A - x and B - y , respectively, by deleting the last entry. T h e entries of each of a and b , therefore, are distinct and lie i n ft - {7}, and thus there exists an element g G G that carries a to b . Since A - x and B - y can be obtained from a a n d b by appending 7, w h i c h is fixed by g , we see that { A - x ) - g = B - y , a n d thus xgy~ is an element of G that carries A to B . T h i s completes the proof.
1 l

We w i l l use L e m m a 8.2 to construct a n infinite family of 3-transitive p e r m u t a t i o n groups, but we start w i t h a family of 2-transitive group actions. R e c a l l that if n > 1 and q is a prime power, t h e n the general linear group G L { n , q ) is the group of invertible n x n matrices over the field F of order q , and the subgroup S L { n , q ) , the special linear group, is the set of determinant-one matrices i n G L { n , q ) . Since G L { n , q ) is isomorphic to the group of invertible linear transformations of an n-dimensional vector space V over F , it is customary to identify the m a t r i x group G L { n , q ) w i t h the corresponding group of linear transformations. One often speaks, therefore, of the "natural action" of G L { n , q ) on V, and the "natural action" of its subgroup S L { n , q ) o n V. B u t , of course, we should remember that the i s o m o r p h i s m between G L { n , q ) a n d the group of linear transformations of V is not really n a t u r a l ; it depends o n the choice of a basis for V. 8 . 3 . L e m m a . L e t V be a n n - d i m e n s i o n a l v e c t o r space o v e r a field F of o r d e r q , w h e r e n > 2. T h e n f o r e v e r y p r i m e - p o w e r q , t h e n a t u r a l a c t i o n of S L { n , q ) o n t h e set of a l l o n e - d i m e n s i o n a l subspaces o f V is doubly transitive.

8A

229

P r o o f . F i r s t , observe that invertible linear transformations of V carry subspaces to subspaces and preserve dimensions, a n d thus S L ( n , q ) actually does permute the set of one-dimensional subspaces of V. A l s o , since n > 2, there are at least two of these, a n d so it makes sense to ask if the action is 2-transitive. N o w let 01,02,61,62 e V be nonzero, and consider the one-dimensional subspaces F o i , F a , F h and F b . A s s u m i n g that F a i + F a a n d t h a t F b i + F b , we must show that there exists a n element g G S L ( n , q ) t h a t carries F a to F b i and F a to F6 . Since F a ^ F a , the vectors a a n d a are linearly independent, a n d so there exist vectors 03, o , . -, a such that { I 1 < i < n } is a basis for V, a n d similarly, we c a n e x p a n d the set {61,62} to a basis { b i I 1 < i < n } .
2 2 2 2 x 2

a i

Since the a; form a basis, there is some (unique) linear transformation t on V t h a t maps a* to bi for 1 < i < n , a n d since the h form a basis, t must be invertible. T h u s t G G L ( n , q ) , a n d t carries F o i to F61 and F a to F6 . T h e proof is not yet complete, however, because t may not have determinant 1.
2

Let 5 = det(t), so that 0 ^ S G F , a n d let e = 8~ . N o w {e&i, 62, ...,&} is a basis for V, and so, reasoning as before, there is a unique invertible linear transformation x on V that carries 61 to e&i and 6 < to 6 ^ for 2 < z < n . T h e m a t r i x corresponding to x w i t h respect to the basis { b i } is diagonal, w i t h (1, l ) - e n t r y equal to e a n d a l l other diagonal entries equal to 1. T h u s det(x) = e, and hence d e t ( i x ) = det(t)det(x) = e5 = 1 and t x G S L ( n , q ) . N o w t x carries 01 to e&i, and so it carries F a to F61. It also carries o to 6 , and hence it carries F a to F6 . T h i s completes the proof.
l x 2

Clearly, scalar m u l t i p l i c a t i o n by a nonzero field element on a vector space fixes a l l subspaces. It follows i n the s i t u a t i o n of L e m m a 8.3 that the subgroup Z of S L ( n , q ) consisting of a l l determinant-one scalar matrices is contained i n the kernel of the action on the set of one-dimensional subspaces of V, a n d thus the factor group S L { n , q ) / Z = P S L { n , q ) acts on this set. It is not h a r d to see t h a t i n fact, Z is the full kernel of the action, and hence P S L { n , q ) acts faithfully, a n d so it is a d o u b l y transitive p e r m u t a t i o n group. W e prove this i n L e m m a 8.29. 8.4. C o r o l l a r y . L e t n > 2. T h e n t h e g r o u p G L { n , 2 ) i s d o u b l y t r a n s i t i v e o n t h e set of n o n z e r o v e c t o r s i n t h e n - d i m e n s i o n a l v e c t o r space V o v e r t h e field of o r d e r 2. P r o o f . Since the field has order 2, a one-dimensional subspace of V contains exactly one nonzero vector, and each nonzero vector, of course, lies i n exactly one one-dimensional subspace. T h e action of G L ( n , 2 ) = S L { n , 2 ) on the

230

8. P e r m u t a t i o n

Groups

nonzero vectors of V, therefore, is permutation-isomorphic to its action on the one-dimensional subspaces of V, w h i c h we know is d o u b l y transitive. N e x t , we need the following easy result. 8 . 5 . L e m m a . Suppose t h a t a g r o u p G acts t r a n s i t i v e l y o n ft, a n d l e t A = G, w h e r e a G ft. T h e n a s u b g r o u p N G i s t r a n s i t i v e o n ft if a n d o n l y if A N = G, a n d N i s r e g u l a r o n ft if a n d o n l y if i n a d d i t i o n , A n N = 1. A l s o , i f N i s r e g u l a r a n d n o r m a l i n G, t h e n t h e c o n j u g a t i o n a c t i o n of A o n t h e n o n i d e n t i t y elements of N i s p e r m u t a t i o n i s o m o r p h i c t o t h e a c t i o n of A o n t h e set ft - { a } .
a

P r o o f . F i r s t , A N = G, if and only if every right coset of A i n G has the form A n for some element n N , and this happens if a n d only if N acts transitively by right m u l t i p l i c a t i o n on the right cosets of A i n G. T h e right m u l t i p l i c a t i o n action of G o n these cosets, however, is p e r m u t a t i o n isomorphic to the action of G on ft, and so N is transitive on the right cosets of A if and only if it is transitive on ft. T h i s shows t h a t as claimed, N is transitive o n ft if a n d only if A N = G. A l s o , assuming that N is transitive, we know t h a t it is regular if and only if the point stabilizer N = A n N is trivial.
a

Now suppose that N is regular and n o r m a l , and let TT : N -> ft be the m a p defined by i r ( n ) = a-n for n G N . T o show t h a t TT is injective, let x,y G N , and suppose i r ( x ) = i r { y ) . T h e n a-x = a-y, and thus x = y since by regularity, x is the unique element of N that carries a to a-x. For surjectivity, observe t h a t since N is transitive, every element of ft has the form a-n = 7r (n) for some element ne N . T h u s TT is a bijection from N to ft, and since TT (1) = a , it follows that vr defines a bijection from N - {1} onto ft - { a } . T o show that this is a p e r m u t a t i o n i s o m o r p h i s m of the actions of A on these two sets, we must check that yr(n ) = yr(n)-a for a G A and nonidentity elements n G N . W e have
a

vr(n ) = a - i a ^ n a ) = ( a - n ) - a = i r ( n ) - a , as required, where the second equality holds because aeA 8.6. C o r o l l a r y . L e t A be a g r o u p t h a t acts v i a g r o u p N , a n d assume t h a t t h e r e s u l t i n g a c t i o n of elements of N i s k - t r a n s i t i v e f o r some p o s i t i v e a n d v i e w N a n d A as s u b g r o u p s of G. T h e n t h e of G o n t h e set of r i g h t cosets of A m G i s ( k + = G.
a

automorphisms o n a finite A o n t h e set of n o n i d e n t i t y i n t e g e r k. L e t G = N x A , right-multiplication action 1)-transitive.

a

P r o o f . L e t ft be the set of right cosets of A i n G, a n d observe that A G , where a G ft is the point a = A . Since G is certainly transitive on ft, it suffices by L e m m a 8.2 to show that A is A;-transitive o n ft - { a } . B u t

8A

231

A N = G and i f l i V = 1, and hence TV is a regular n o r m a l subgroup of G by L e m m a 8.5. A l s o by L e m m a 8.5, the a c t i o n of A on ft - { a } is p e r m u t a t i o n isomorphic to its action on N - {1}, a n d by hypothesis, this latter action is /c-transitive. T h i s completes the proof. F i n a l l y , we c a n construct the promised family of 3-transitive p e r m u t a t i o n groups. 8.7. C o r o l l a r y . L e t V be a n e l e m e n t a r y a b e l i a n 2 - g r o u p of o r d e r 2 w i t h n > 2 , a n d w r i t e G = V x G L ( n , 2 ) , w h e r e t h e a c t i o n of G L ( n , 2 ) i s d e f i n e d by i d e n t i f y i n g V w i t h a v e c t o r space of d i m e n s i o n n o v e r a field of o r d e r 2. T h e n G i s a 3 - t r a n s i t i v e p e r m u t a t i o n g r o u p of d e g r e e 2 .
n n

P r o o f . W r i t e A = G L ( n , 2 ) . B y C o r o l l a r y 8.4, the action of A o n the nonidentity elements of V is 2-transitive, a n d hence by C o r o l l a r y 8.6, the a c t i o n of G on the right cosets of A is 3-transitive. T h e degree of this action is \ G : A \ = \V\ = 2 , and so a l l t h a t remains is to show that the action of G on the right cosets of A is faithful. T h e kernel K of the action, however, is n o r m a l i n G, and since K C A , we have V n K = 1, a n d so K C C ( V ) . T h e n K acts t r i v i a l l y on V, and since the n a t u r a l action of A = G L ( n , 2) on V is faithful, it follows that K = 1.
n G

A c t u a l l y , C o r o l l a r y 8.6 is less general t h a n it m a y appear to be. In fact, the integer k i n the statement of that result c a n almost never exceed 2. 8.8. L e m m a . L e t A be a g r o u p t h a t acts v i a a u t o m o r p h i s m s o n a finite N , a n d suppose t h a t t h e r e s u l t i n g a c t i o n of A o n t h e n o n i d e n t i t y o f N is k-transitive, with k > \ . Then (a) N i s a n e l e m e n t a r y a b e l i a n p - g r o u p f o r some prime p. group elements

(b) I f k > 1 t h e n e i t h e r p = 2 o r \ N \ = 3. (c) I f k > 2 , t h e n \ N \ = 4 a n d k = 3.

P r o o f . L e t x G N have prime order p . Since A acts v i a automorphisms and is transitive o n the nonidentity elements of N , it follows that every nonidentity element of N has order p , a n d hence TV is a p-group. T h e n Z(7Y) is n o n t r i v i a l , and so we can assume t h a t x is central i n N , and hence every nonidentity element of N is central. T h u s N is abelian, p r o v i n g (a). Now assume that the a c t i o n of A o n the nonidentity elements of N is 2-transitive. If p > 2 then x a n d x ' are distinct nonidentity elements, and if |TV| > 3, we can choose a nonidentity element y of N different from b o t h x and x ~ . B y 2-transitivity, there is an element a G A t h a t carries the ordered pair { x , x ~ ) to the ordered pair ( x , y ) . T h e n x = x and y = ( a T ) " = ( x - ) ' = x ~ . T h i s c o n t r a d i c t i o n proves (b).
1 l l a 1 0 1 l

232

8. P e r m u t a t i o n

Groups

F i n a l l y , assume that the action is 3-transitive. I n particular, 3 < k < \ N - {1}|, so \ N \ > 4, a n d we have p = 2 by (b). Choose a nonidentity element y of N different from x , a n d observe that xy + 1 a n d that xy is different from b o t h x a n d y . i f |7Y| > 4, choose a nonidentity element z different from x , y and xy, and consider the ordered triples ( x , y , x y ) and { x , y , z ) , w h i c h consist of distinct elements. B y 3-transitivity, there exists a G A such t h a t x = x , y = y and z = { x y ) = x y = xy, w h i c h is a contradiction. T h u s |7V| = 4, and so k < \ N - {1}| = 3.
a a a a a

We close this section w i t h a v a r i a t i o n o n L e m m a 8.8(a). In order to conclude that N is a n elementary abelian p-group i n t h a t lemma, we shall see t h a t the assumption t h a t A acts transitively on the set of nonidentity elements of N is unnecessarily strong. We say that an action of G on is h a l f - t r a n s i t i v e if a l l orbits have equal size. ( A transitive action, therefore, is certainly half-transitive.) T h e following is a result of D . S. P a s s m a n and the author. 8 . 9 . T h e o r e m . L e t A be a finite g r o u p t h a t acts f a i t h f u l l y v i a phisms o n a finite g r o u p N , a n d assume that theresulting action t h e n o n i d e n t i t y elements of N i s h a l f - t r a n s i t i v e . T h e n e i t h e r t h e F r o b e n i u s , o r else N i s a n e l e m e n t a r y a b e l i a n p - g r o u p f o r some a n d n o n o n i d e n t i t y p r o p e r s u b g r o u p of N a d m i t s t h e a c t i o n of A . automorof A o n action is prime p ,

N o t e that a Frobenius action of A o n N is a u t o m a t i c a l l y half-transitive on the nonidentity elements of N because a l l A - o r b i t s o n this set have equal size \ A \ . We need a p r e l i m i n a r y result about finite groups p a r t i t i o n e d by n o r m a l subgroups. 8 . 1 0 . L e m m a . L e t G be a finite g r o u p a n d l e t U be a c o l l e c t i o n of p r o p e r n o r m a l s u b g r o u p s o f G . Suppose t h a t G = {JTL a n d t h a t d i s t i n c t members o/n i n t e r s e c t t r i v i a l l y . T h e n G i s a n e l e m e n t a r y a b e l i a n p - g r o u p f o r some prime p. P r o o f . L e t X G II, and observe that i f X ^ Y G II, then X n Y = 1, and so Y C C ( X ) . It follows that
G

G = XU

|J
yen

y c i u CQ(X) .

But G cannot be the u n i o n of two proper subgroups, a n d since X is proper, it follows t h a t C ( X ) = G, and hence X C Z ( G ) . Since | J II = G and every member of II is central, we have Z ( G ) = G , and thus G is abelian.
G

8A

233

If x , y G G have different orders, we argue t h a t they lie i n a c o m m o n member of II. T o see this, assume t h a t o ( x ) > o(y). T h e n ( )(y)
xy

o i x

- y

y )

o { y )

= x

o ( y )

1,

where the first equality is v a l i d because G is abelian a n d the inequality holds because o { y ) < o { x ) . N o w let X , Z II, w i t h x G X a n d xy G Z . T h e n x( ) = {xy)^ is a nonidentiy element i n X n Z , a n d thus X = Z . It follows t h a t xy G X , a n d so y X , as claimed.
y

Let x G G have p r i m e order p, a n d let X G II w i t h W e complete the proof now by showing that every nonidentity element y G G has order p . If y G G - X , then s and y do not lie i n a c o m m o n member of II, a n d so o(y) = o ( x ) = p , as wanted. If, o n the other h a n d , y is a nonidentity element of X , choose z G G - X . T h e n 2 does not share a member of II w i t h either x or y, a n d thus o(y) = o { z ) = o { x ) = p . P r o o f o f T h e o r e m 8.9. For nonidentity elements x G N , w r i t e G = C ( C { x ) ) , and let II = { C | 1 / 1 JV} be the collection of subgroups of N constructed i n this way. It is clear t h a t x G C , and it follows that (jn = N . If 1 ^ y G C , we argue that G ^ = G . T o see this, observe that C { x ) centralizes y, and thus C ( x ) C C ^ ( y ) . B u t C U ( x ) a n d C ( y ) are subgroups of index r i n A , where r is the c o m m o n size of the A - o r b i t s on N - {1}, and thus equality holds, a n d C { x ) = C { y ) . T h e n
x N A x x x x A A A A A

C as claimed.

= C {C (y))
N A

= C {C (x))
N A

= C ,
x

I n p a r t i c u l a r , this shows that distinct members of the set II

intersect t r i v i a l l y . Now assume t h a t the action of A o n TV is not Frobenius, so that r < \ A \ . T h e n C { x ) > 1 for nonidentity elements x G TV, a n d since the a c t i o n of A on N is faithful, C { x ) does not act t r i v i a l l y on N . I n other words, C < N , and so II consists of proper subgroups of N .
A A x

We w o r k now to show that the members of II are n o r m a l i n N so that we can a p p l y L e m m a 8.10. F i r s t , observe t h a t r divides |7V| - 1, and thus r a n d \ N \ are coprime. L e t { x i , x , . . . , x } be a n A - o r b i t of nonidentity elements of N , a n d let K be the conjugacy class of i n N . (Note t h a t perhaps the classes K are not a l l distinct.)
2 r t X i t

Since A acts v i a automorphisms o n N a n d A permutes the elements X i transitively, it follows t h a t A permutes the set of conjugacy classes K transitively. In p a r t i c u l a r , these classes have equal cardinality, say k, where k is a divisor of \ N \ , a n d hence k is relatively p r i m e to r .
{

N o w write
r

234

8. P e r m u t a t i o n

Groups

a n d recall t h a t \ K i \ = k for a l l i. Since the sets K a r e conjugacy classes of N , distinct K i are disjoint, a n d it follows t h a t \ U \ is a m u l t i p l e of k. Furthermore, since the sets K i are p e r m u t e d by A , their u n i o n U is an A invariant subset of N . A l s o , U consists of nonidentity elements, and so it is a u n i o n of A - o r b i t s of size r , and we deduce that r divides \ U \ . N o w k and r are coprime integers and b o t h d i v i d e \ U \ , and it follows that r k divides \ U \ , and i n particular, r k < \ U \ . B u t U the u n i o n of the sets K i for 1 < i < r , a n d each of these sets has c a r d i n a l i t y k, and thus \ U \ < r k . T h u s \ U \ = r k , and so the sets K i must a l l be different. T h i s shows t h a t a nonidentity class of N cannot contain distinct members of an A - o r b i t .
{

We argue n e x t t h a t every nonidentity conjugacy class of N is contained in one of the subgroups C i n IT Since distinct members of n intersect trivially, this w i l l show that each member of II is a u n i o n of classes of N , a n d thus is n o r m a l i n N , as wanted. G i v e n a nonidentity class K of N , let x G K . Since A permutes the classes of N and C { x ) fixes x , it follows t h a t C { x ) stabilizes the class AT, a n d thus C ( x ) permutes the elements of K . W e know, however, t h a t distinct members of K cannot lie i n a c o m m o n A - o r b i t , and this shows that, i n fact, C { x ) fixes every member of K . T h u s K C C ( C { x ) ) = C , as claimed, and hence we can apply L e m m a 8.10 to conclude t h a t N is an elementary abelian p-group for some p r i m e p .
x A A A A N A x

F i n a l l y , if H is a proper A - i n v a r i a n t subgroup of TV, we work to show t h a t H = 1. L e t { x i , x , . . . , x } be an A - o r b i t outside of H , and observe t h a t A permutes the set of cosets H x and so the set
2 r u

V = \ j H x i
i=i

is A - i n v a r i a n t . Since V C G - H , it consists of nonidentity elements, a n d so it is a u n i o n of A - o r b i t s of size r . T h u s r divides \V\, a n d also, since V is a u n i o n of cosets of H , we see t h a t \ H \ divides | V | . B y coprimeness, r \ H \ divides | V | , a n d so | V | > r \ H \ , and it follows t h a t the cosets H are distinct. T h i s shows t h a t no right coset of H other t h a n H can contain distinct elements of an A - o r b i t .
X i

If H > 1, then H can be contained i n at most one member of II. W e w i l l show that, i n fact, H is contained i n some member C G II, a n d that C also contains every element x G N - H . It follows that C = N , which is a contradiction, and this w i l l prove that H = 1, as wanted. It suffices, therefore, to show that if x G N - H , then H C C .
x

Since A permutes the right cosets of H , it follows that C { x ) stabilizes H x , a n d so C { x ) permutes the elements of H x . B u t H x cannot contain distinct elements of an A - o r b i t , and thus C { x ) fixes every element of H x .
A A A

Problems 8 A

235

W e deduce that H x C C , and since C is a subgroup t h a t contains x , we conclude that H C C . A s we have seen, this completes the proof.
x x x

N o t e t h a t i n the exceptional case of T h e o r e m 8.9, where the action of A on TV is Frobenius, we can conclude t h a t N is nilpotent by the theorem of T h o m p s o n ' s thesis. T h u s , i n a l l cases, i f A acts n o n t r i v i a l l y o n N a n d half-transitively on the nonidentity elements, then N is nilpotent. Problems 8A 8 A . 1 . If A and B are nonisomorphic groups w i t h | A | = show that there exists a p e r m u t a t i o n group G t h a t has regular subgroups isomorphic to each of A and B . A l s o , decide whether or not a p e r m u t a t i o n group can have nonisomorphic regular n o r m a l subgroups. H i n t . F o r the first part, take G to be a n appropriate s y m m e t r i c group. 8 A . 2 . L e t H be a transitive subgroup of some p e r m u t a t i o n group G. Show t h a t C { H ) is semiregular. Deduce t h a t a n abelian transitive p e r m u t a t i o n group must regular.
G

8 A . 3 . G i v e n a finite group G w i t h Z ( G ) = 1, show t h a t there exists a p e r m u t a t i o n group t h a t has two distinct regular n o r m a l subgroups, each isomorphic to G. 8 A . 4 . Suppose t h a t U and V a r e regular n o r m a l subgroups of some permut a t i o n group G. If U n V = 1, show t h a t U a n d V are isomorphic and have t r i v i a l centers. H i n t . It is no loss to assume G = U V . 8 A . 5 . L e t G act fc-transitively on some set ft, a n d let H = G , where a ft. L e t A be the set of points i n ft t h a t are fixed by H . Show t h a t N ( 7 f ) acts r - t r a n s i t i v e l y on A , where r is the m i n i m u m of k a n d | A | .
a G

8 A . 6 . L e t G act transitively on ft, and let H = G , where a ft. L e t P G S y l ( / f ) , and let A be the set of points i n ft t h a t are fixed by P . Show t h a t N ( P ) acts transitively o n A .
a p G

8 A . 7 . Suppose t h a t a n abelian group A acts faithfully o n a group N , a n d assume t h a t its a c t i o n on the nonidentity elements of N is half-transitive. Show that the action is Frobenius, a n d deduce t h a t A is cyclic. 8 A . 8 . L e t G act transitively o n ft, a n d suppose t h a t N < G . Show t h a t G permutes the A - o r b i t s i n ft transitively, a n d deduce t h a t N is half-transitive on ft.

236

8. P e r m u t a t i o n

Groups

8 A . 9 . L e t G act 2-transitively on ft, a n d suppose t h a t N < G and that N acts nontrivially. Show t h a t N acts transitively. N o t e . I n fact, 2-transitivity is more t h a n is really needed here. T h e right c o n d i t i o n is p r i m i t i v i t y , w h i c h we w i l l discuss i n the next section. 8 A . 10. If G is a solvable 4-transitive p e r m u t a t i o n group, show that G is isomorphic to the s y m m e t r i c group 5 .
4

H i n t . Show t h a t a m i n i m a l n o r m a l subgroup N of G is regular, and consider the conjugation action of a point stabilizer G on N .
a

N o t e . I n fact, 5 and S are the only solvable 3-transitive p e r m u t a t i o n groups, but this is somewhat harder to prove.
4 3

8 A . 11. Show t h a t for every p r i m e power q > 1, there exists a solvable sharply 2-transitive p e r m u t a t i o n group of degree q. H i n t . Consider a field of order q. 8 A . 12. L e t G act transitively o n ft, a n d let t a t i o n character. (Recall that by definition, formula ( g ) = \ { a G ft | a - g = a } \ . ) Show and o n l y if the average value of { g ) for g G
X X X 2 X

be the corresponding permuis the function defined by the that G is 2-transitive o n ft if G is 2.

2 X

H i n t . F i n d an action of G whose p e r m u t a t i o n character is , and recall that, i n general, the average value of a p e r m u t a t i o n character is equal to the number of orbits. (See P r o b l e m 1 A . 6 and the note following it.) 8 A . 1 3 . L e t G act 2-transitively on ft, and let be the corresponding perm u t a t i o n character. F i n d a positive integer m such t h a t G is 3-transitive o n ft if and o n l y i f the average value of ( g ) for g G G is m .
X 3 X

8 A . 1 4 . L e t G act transitively on ft, and suppose H i s a subgroup of G w i t h \ G : H \ = m . Show t h a t H has at most m orbits on ft, a n d t h a t i f H contains no point stabilizer i n G, then H has at most m/2 orbits o n ft. H i n t . W h a t is the smallest possible size of a n i f - o r b i t ? 8 A . 1 5 . If H C G, we c a n define a n action of H x H o n G by setting g - ( x , y ) = x ~ g y for g G G a n d x,y G H . Show that the action of G on the right cosets of H is 2-transitive if and o n l y i f H x H has exactly two orbits on G.
l

8 A . 1 6 . L e t G be 2-transitive on a set ft, where |ft| = n . Suppose that H C G is transitive, a n d t h a t \ G : H \ is relatively p r i m e to n - 1. Show that H is 2-transitive.

8B

237

8 A . 1 7 . L e t q be a power of 2, a n d let S L ( 2 , q ) act on the q + 1 onedimensional subspaces of a two-dimensional vector space over a field of order q. Show t h a t this action is sharply 3-transitive.

8B Suppose t h a t G acts transitively on ft, a n d let A C ft be a nonempty subset. If g G G, we w r i t e A - g = { a - g \ a G A } , and we refer to A - g as a translate of A . T h e nonempty set A is said to be a block i f every one of its translates other t h a n A itself is disjoint from A . I n other words, A is a block if a n d only if A n A - g = 0 whenever g G G a n d A - g + A . W e stress that we speak of "blocks" only for transitive actions, and that, by definition, blocks are nonempty. For example, consider the group G of symmetries of square A B C D , and view G as a (transitive) p e r m u t a t i o n group on the vertex set ft = { A , B , C, D } . (Note t h a t G = D , a n d , i n fact, G is a full Sylow 2-subgroup of the s y m m e t r i c group on ft.) N o w choose a diagonal of the square, say A C , a n d let A = { A , C } , the set consisting of the ends of that diagonal. E a c h s y m m e t r y g G G carries the diagonal A C to a diagonal, and so the only translates of A are A itself and the set { B , D } of endpoints of the other diagonal. In p a r t i c u l a r , if A - g ^ A , then A - g n A is empty, and so A is a block.
8

For a more general example, suppose H C G, a n d let ft = { H g \ g e G } , so t h a t as usual, G acts transitively by right m u l t i p l i c a t i o n on ft. Suppose t h a t H C K C G, where K is a subgroup, and let A = { H k \ k G K } . T h e n A is a nonempty subset of ft, and i f g G G, it is easy to see that the translate A - g is exactly the set of those right cosets of H i n G t h a t happen to be contained i n the coset K g of K . If K g = K , t h e n A - g = A , a n d if K g ^ K , then K g n K = 0 , and so no member of A - g c a n be a member of A . W e have A - g D A = 0 i n this case, a n d so A is a block. Of course, i n an a r b i t r a r y transitive action, the whole set ft is a block, and so is every one-point subset of ft. These are referred to as t r i v i a l blocks, and we say t h a t the given transitive a c t i o n of G on ft is p r i m i t i v e i f the t r i v i a l blocks are the only blocks; otherwise, the a c t i o n is i m p r i m i t i v e . T h u s , for example, the action of the group of symmetries of a square on the four vertices is i m p r i m i t i v e , a n d the a c t i o n of a group G on the right cosets of a subgroup H is i m p r i m i t i v e if there exists a subgroup K such t h a t H < K < G. (The c o n d i t i o n H < K guarantees that the block A = { H k | k G K } contains more t h a n one point, a n d the c o n d i t i o n K < G guarantees t h a t A is not a l l of ft.)

238

8. P e r m u t a t i o n

Groups

8.11. L e m m a . L e t G be a g r o u p t h a t acts t r a n s i t i v e l y o n a set ft, a n d suppose t h a t A C ft i s a b l o c k . T h e n | A | d i v i d e s |ft| a n d A has e x a c t l y | f t | / | A | dzsimci t r a n s l a t e s . These translates a r e disjoint blocks; t h e i r u n i o n i s ft, a n d they a r e t r a n s i t i v e l y p e r m u t e d by G. P r o o f . L e t a,b G G, and suppose that A - a + A - b . A p p l y b~ to deduce t h a t A - i a b ' ) + A , and thus A ^ a b ' ) n A = 0 since A is a block. N o w apply 6 to o b t a i n A - a n A - b = 0. T h i s shows that the distinct translates of A are a l l disjoint, and, i n particular, a l l translates of A are blocks. O f course, G permutes the set of translates of A transitively.
l 1 1

T h e u n i o n of the distinct translates of A is a nonempty G-invariant subset of ft, a n d since G is transitive on ft, this u n i o n must be a l l of ft. Since the distinct translates of A are disjoint, have equal c a r d i n a l i t y and cover ft, there must be exactly | f t | / | A | of them, and so this number is an integer. 8.12. C o r o l l a r y . A t r a n s i t i v e g r o u p a c t i o n o n a set of p r i m e c a r d i n a l i t y i s primitive. P r o o f . L e t ft be the relevant set, where |ft| is prime. If A C ft is a block, t h e n | A | divides |ft|, and so | A | = |ft| or | A | = 1, and hence either A = ft or A consists of a single point. In either case, A is a t r i v i a l block, and thus the given a c t i o n is p r i m i t i v e . T h e following l e m m a gives a general, and fairly complete description of the blocks of a transitive action, and it leads to a useful necessary and sufficient c o n d i t i o n for p r i m i t i v i t y . T o state our result, we introduce a bit of notation. If G acts on ft and a G ft, then for each subgroup AT C G, we write a - K = { a - k | k G K } . In other words, a-AT is the AT-orbit containing the point a . 8.13. L e m m a . L e t G be a g r o u p t h a t acts t r a n s i t i v e l y o n a set ft, a n d l e t H = G , w h e r e a G ft. / / AT D H i s a s u b g r o u p , t h e n a - K i s a b l o c k c o n t a i n i n g a a n d \ a - K \ = \ K : H \ . A l s o , t h e m a p K i-> a - K i s a b i s e c t i o n f r o m t h e set of s u b g r o u p s { K | H C K C G } o n t o t h e set of a l l b l o c k s t h a t contain a .
a

P r o o f . Suppose that g G G a n d that the sets a - K a n d ( a - K ) - g are not disjoint. W e argue t h a t g G K . To see this, observe that there exist elements x , y e K such t h a t a-x = ( a - y ) - g . T h e n ygx' G G = H C AT, and since x a n d y are i n K , it follows t h a t g G AT, as claimed.
1 a

T h e set a - K is certainly AT-invariant, and the result of the previous paragraph shows t h a t K is the full stabilizer of this set i n G. T h u s a - K uniquely determines AT, and it follows t h a t the m a p K H-> a - K is injective

8B

239

on the set { K \ H C K C G } . T h e result of the previous paragraph also shows t h a t the set A = a - K is a block. T o see this, observe t h a t if A is not disjoint from A - g , t h e n g G K , and thus A - g = A . Furthermore, the block A certainly contains a , and since it is the Af-orbit containing a , we have | A | = \ K : H \ by the fundamental c o u n t i n g p r i n c i p l e . We must now show t h a t given a n a r b i t r a r y block A containing a , there exists a subgroup K D H such t h a t A = a - K . L e t K be the stabilizer of A i n the action of G on subsets of ft, and observe that a - K C A since K stabilizes A and a G A . T o prove the reverse containment, let 5 G A . Since G is transitive o n ft, we can choose g G G w i t h a - g = S. T h e n 8 G A n A - g , a n d thus A = A - g because A is a block. T h u s g e K , a n d 5 = a - g G a - K , a n d this shows that A = a - K , as required. A l s o , we c a n take S = a and g e H m the above argument, and since we showed t h a t g K , it follows t h a t H C K . T h i s completes the proof. 8 . 1 4 . C o r o l l a r y . L e t G be a g r o u p t h a t acts t r a n s i t i v e l y o n a set ft c o n t a i n i n g m o r e t h a n one p o i n t , a n d l e t H = G , w h e r e a G ft. T h e n G i s p r i m i t i v e o n ft if a n d o n l y if H i s a m a x i m a l s u b g r o u p o f G .
a

P r o o f . F i r s t , assume that H is m a x i m a l . T h e n by L e m m a 8.13, the only blocks t h a t contain a are a - H = { a } a n d a - G = ft. If A is a n a r b i t r a r y block, then some translate A - g of A contains a , and A - g is a block by L e m m a 8.11. It follows that A - g is either { a } or ft, and thus A is a t r i v i a l block. T h i s shows that G is p r i m i t i v e on ft. Conversely, i f G is p r i m i t i v e , then a l l blocks containing a are t r i v i a l , a n d thus there are precisely two such blocks: ft a n d { a } . B y L e m m a 8.13, therefore, there are exactly two subgroups of G t h a t contain H , and thus H is m a x i m a l i n G. N o t e t h a t C o r o l l a r y 8.14 renders C o r o l l a r y 8.12 superfluous since i f |ft| is prime, then a point stabilizer has p r i m e index, a n d hence it is a m a x i m a l subgroup. It follows by 8.14 that G is p r i m i t i v e . The following result gives a useful connection between p r i m i t i v i t y and n o r m a l subgroups. 8 . 1 5 . L e m m a . L e t G be a g r o u p t h a t acts t r a n s i t i v e l y o n a set ft, a n d l e t N < G. T h e n each o r b i t of t h e a c t i o n of N o n ft i s a b l o c k f o r G. I n p a r t i c u l a r , if G i s p r i m i t i v e o n ft, t h e n t h e a c t i o n of N i s e i t h e r t r i v i a l o r transitive. P r o o f . L e t g G G, and observe t h a t if A is an orbit for the action of N , then A - g is an orbit for the action of N = N . ( T h i s is immediate from the fact t h a t if a-n = (5 w i t h a , (5 G ft and n G N , t h e n ( a - g ) - ( n ) = p - g . ) B u t distinct TV -orbits i n ft are disjoint, a n d thus if A - g + A , then A - g n A = 0 .
9 9

240

8. P e r m u t a t i o n

Groups

In other words, A is a block for G, as wanted. If G is p r i m i t i v e , it follows t h a t either A = ft, or else | A | = 1. In other words, if TV is not transitive on ft, t h e n a l l TV-orbits have size 1, and thus TV acts t r i v i a l l y . A n alternative argument for the final assertion of L e m m a 8.15 is as follows. If the action of G is p r i m i t i v e a n d A is a point stabilizer, then H is a m a x i m a l subgroup of G by C o r o l l a r y 8.14. Since H C H N C G, there are just two possibilities. E i t h e r H N = G, i n w h i c h case it is easy to see t h a t TV is transitive o n ft, or else H N = H . I n the latter situation, TV C H , and thus TV is contained i n every conjugate of H , and it follows t h a t TV fixes a l l points i n ft.
7

N e x t , we relate p r i m i t i v i t y and m u l t i p l e t r a n s i t i v i t y . 8.16. L e m m a . Suppose that G is a group w i t h a doubly transitive o n ft. T h e n G i s p r i m i t i v e o n ft. action

P r o o f . If G is i m p r i m i t i v e , there exists a n o n t r i v i a l block A , a n d since A does not consist of just one point, we can choose distinct points a , 0 G A . A l s o , A is not a l l of ft, and so we can choose 7 G ft A , and, i n particular, 7 ^ a . B y 2-transitivity, there exists an element gG t h a t carries the pair (a,/3) to the pair ( a , 7). I n particular, a - g = a , and thus a G A n A - g . But A is a block, a n d this forces A - g = A , a n d since 0 G A , we have 0-g G A - g = A . T h i s is a contradiction, however, because 0-g = 7 and 7 0 A. It is easy to find examples of p r i m i t i v e actions t h a t are not 2-transitive. For example, consider the d i h e d r a l group D of order 2 p , where p > 3 is prime. B y C o r o l l a r y 8.12 or C o r o l l a r y 8.14, this group acts p r i m i t i v e l y o n the p cosets of a subgroup of order 2. B u t p - 1 does not d i v i d e | D | , so the a c t i o n cannot be 2-transitive. I n this context, we m e n t i o n a remarkable result of W . B u r n s i d e t h a t shows t h a t a (necessarily p r i m i t i v e ) transitive p e r m u t a t i o n group of prime degree p must be 2-transitive except possibly w h e n a Sylow p-subgroup is n o r m a l . B u r n s i d e ' s proof of this fact was character theoretic, but a c o m p a r a t i v e l y easy non-character proof was found by H . W i e l a n d t . ( W i e l a n d t ' s proof can be found i n the b o o k P e r m u t a t i o n G r o u p s by D . S. Passman.)
2 p 2 p

A s B u r n s i d e ' s theorem o n p r i m i t i v e p e r m u t a t i o n groups of prime degree suggests, once one knows that a group acts p r i m i t i v e l y o n some set, it sometimes does not require a great deal of a d d i t i o n a l i n f o r m a t i o n to deduce that the action is m u l t i p l y transitive. F o r example, suppose t h a t G is a p r i m i t i v e p e r m u t a t i o n group on ft, and assume that G contains at least one transposition. (Recall t h a t a t r a n s p o s i t i o n is a p e r m u t a t i o n t h a t exchanges two

8B

241

points and fixes a l l others. A l s o , recall the n o t a t i o n ( a , (5) for the transposition exchanging a w i t h /?.) If G is p r i m i t i v e a n d contains a transposition, a theorem of C . J o r d a n asserts t h a t G is the full s y m m e t r i c group o n ft, and so, of course, it is m u l t i p l y transitive. W e w i l l o b t a i n this result as a corollary of a m u c h more general theorem, also due to J o r d a n , b u t there is also an entirely elementary a n d rather geometric proof t h a t we cannot resist presenting first. (We also cannot resist m e n t i o n i n g t h a t we are following i n J o r d a n ' s footsteps here: he was the author of the very first group-theory book, p u b l i s h e d i n 1870. H i s Trait des S u b s t i t u t i o n s , of course, deals only w i t h p e r m u t a t i o n groups since C a y l e y ' s m o d e r n , abstract point o f view was not enunciated u n t i l 1878.) 8 . 1 7 . T h e o r e m ( J o r d a n ) . L e t G be a p r i m i t i v e p e r m u t a t i o n g r o u p o n ft, a n d suppose that G contains a transposition. T h e n G i s the full s y m m e t r i c g r o u p o n ft.

P r o o f . Since every element of Sym(ft) is a p r o d u c t of transpositions, it suffices to show t h a t G contains a l l transpositions o n ft. T o do this, consider the (undirected) graph w i t h vertex set ft, i n w h i c h distinct points a a n d p are j o i n e d by an edge precisely w h e n the t r a n s p o s i t i o n ( a , 3) is a n element of G. If x = { a , 3) a n d g G G is arbitrary, it is easy to see t h a t x is the transposition ( a - g , p - g ) , w h i c h exchanges a - g and p - g . O f course, i f x G G, then also x G G, and this shows t h a t i f a a n d P are two points j o i n e d by an edge i n our graph, then their images under g are also j o i n e d b y a n edge. In other words, the permutations i n G define automorphisms of the graph; they carry points to points and edges to edges.
9 9

Now consider the connected components of the graph, viewing each component as a n o n e m p t y set of vertices. (In other words, no edge joins a point in one component to a point i n a different component, a n d w i t h i n each component, any two distinct points can be j o i n e d b y a p a t h consisting of one or more edges.) If A is such a component, t h e n since g G G induces a g r a p h a u t o m o r p h i s m , it follows t h a t A - g must also be a component. N o w i f A meets A - g nontrivially, then A U A - g is a connected set that contains the components A and A - g . T h i s set, therefore must be equal to each of these components, a n d thus A = A - g . I n other words, i f A a n d its translate A - g are not disjoint, they cannot be distinct, a n d hence A is a block. T h e a c t i o n of G is p r i m i t i v e , however, a n d hence A must be a t r i v i a l block, a n d so either A is a l l of ft or A consists of just one point. B y hypothesis, our g r a p h has at least one edge, and so there is a component A consisting of more t h a n one point. T h e n A = ft, w h i c h means t h a t the graph is connected.

242

8. P e r m u t a t i o n

Groups

E a c h two points a , /? G ft are joined by a p a t h , and we define the distance d ( a , p ) to be the length of the shortest p a t h from a to p , where we set d ( a , a ) = 0. T h u s d ( a , /?) = 1 precisely when a ^ /3 and there is an edge j o i n i n g a to /?. E q u i v a l e n t ^ , d ( a , 0 ) = 1 if a n d o n l y i f the group G contains the transposition ( a , P ) . It suffices, therefore, to show that d ( a , p ) < 1 for all a , P e n . Suppose (for a contradiction) t h a t there exist a a n d 0 i n ft w i t h d = d { a , P ) > 1, a n d choose these points so that d > 1 is as s m a l l as possible. Since there is a p a t h of length d j o i n i n g a and p , there must exist a point 7 G ft such t h a t d ( a , 7) = d - 1 a n d d ( , 0 ) = 1. N o w d - 1 > 1, and so by the m i n i m a l i t y of d, it follows t h a t d ( a , ) = 1, and thus the transposition x = ( a , 7) is an element of G , as is the transposition y = (7,/?). T h e n f = (a-y, 7-y) = ( a , /?), and of course x G G . T h e n a and /? are j o i n e d by an edge, a n d this is a c o n t r a d i c t i o n since d ( a , P ) = d > 1.
7 7 y x

To write

state the more general theorem of J o r d a n , we introduce some adIf G acts on ft and A C ft is a n a r b i t r a r y subset, we
G =

d i t i o n a l notation.

{9

G G I a - g = a for a l l a G A } ,

so t h a t G

is the pointwise stabilizer of A . T h i s should be distinguished

= { g G G I A - g = A} .
A

from the setwise stabilizer,

G
(
( A )

N o t e t h a t G A ) acts on A , and G G
A

is the kernel of this action, so that

( Q

< G A).
(

A l s o , note t h a t G

( A )

= G

A )

, and thus G Q _

<G

( A )

8.18. T h e o r e m (Jordan). L e t G be a g r o u p w i t h a p r i m i t i v e a c t i o n o n a set ft, a n d l e t A C ft w i t h | A | < |ft| - 2. Suppose i / i a i G acis p r i m i t i v e l y o n ft - A . T a e n ifte a c i i o n of G o n Q i s ( | A | + l ) - i r a r w t i t e .

A

To see t h a t T h e o r e m 8.18 actually does i m p l y T h e o r e m 8.17, suppose t h a t G is a p r i m i t i v e p e r m u t a t i o n group on ft, and assume t h a t the transp o s i t i o n ( a , P) is an element of G . Take A = ft - { a , P } , so t h a t | A | = n - 2, where n = |ft|. Since ( a , P ) G G , it follows t h a t G is transitive on { a , P } = Q - A , and, i n fact, this action is p r i m i t i v e since ft - A has prime c a r d i n a l i t y 2. B y T h e o r e m 8.18, therefore, G is (n - Intransitive on ft, and so
A A

| G | > | 0 _ i ( f t ) | = n - ( n - 1)
n

3-2 = n ! = |Sym(ft)| .

T h u s G is the full s y m m e t r i c group on ft, as is asserted by T h e o r e m 8.17. We also have the following closely related result.

8.19. C o r o l l a r y . Suppose t h a t G i s a p r i m i t i v e p e r m u t a t i o n g r o u p o n ft, a n d assume t h a t G c o n t a i n s a 2,-cycle. T h e n G i s e i t h e r the f u l l s y m m e t r i c g r o u p o r t h e a l t e r n a t i n g g r o u p o n ft.

8B

243

P r o o f . Suppose that ( 0 , ^ , 7 ) is a 3-cycle contained i n G, and let A = ft - {a,/3,7>. T h e n G contains the 3-cycle ( a , , 7 ) , a n d hence G is transitive on ft - A , and it is p r i m i t i v e since 3 is prime. W r i t i n g n = |ft|, we have | A | + 1 = n - 2, and it follows b y T h e o r e m 8.18 that G is (n - 2)transitive on ft. A s we have seen, this guarantees t h a t G is either the full s y m m e t r i c group or the alternating group on ft.
A A

W e begin w o r k t o w a r d a proof of T h e o r e m 8.18 w i t h the following sufficient c o n d i t i o n for p r i m i t i v i t y . 8 . 2 0 . L e m m a . L e t G be a g r o u p a c t i n g t r a n s i t i v e l y o n ft, a n d l e t H G be a s u b g r o u p . L e t X ft be a n H - o r b i t , a n d assume that H is primitive o n X a n d t h a t \X\ > | f t | / 2 . T h e n G i s p r i m i t i v e o n ft. P r o o f . L e t A be a block for the action of G o n ft. W e argue t h a t if A < ft, then A n X contains at most one point. T o see this, let h G H , and observe t h a t ( A n X ) - h = A - h n X . If this set is different from A n X , then clearly A - h ^ A , and since A is a block for G, it follows t h a t A - h n A = 0. If ( A n X ) - h ^ ( A n l ) , therefore, we have ( A n X ) n ( A n X ) - h = 0, and so if A n X is nonempty, it is a block for the a c t i o n of H on X . B y assumption, however, H is p r i m i t i v e on X , and thus i f A n A is nonempty, it must either be a l l of X , or else it contains just one p o i n t . If A n X = X , then A D A , and so | A | > | A | > | f t | / 2 . B y L e m m a 8.11, however, | A | is a divisor of |ft|, and hence i n this case, A = ft, w h i c h is contrary to assumption. It follows that | A n A I < 1, as claimed. N o w suppose t h a t A is an a r b i t r a r y block for the action of G on ft. We must show t h a t A is a t r i v i a l block, and so assuming that A < ft, our goal is to show that | A | = 1. B y L e m m a 8.11, the translates of A are blocks, and, of course, none of t h e m is a l l of ft. T h e r e are exactly | f t | / | A | such translates, and every point of ft is i n one of t h e m . B y the result of the previous paragraph, each translate of A contains at most one point of X , a n d thus the number of translates is at least equal to \X\. T h e n | f t | / | A | > \X\ > | f t | / 2 , and hence | A | < 2. T h u s | A | = 1, as required. Suppose t h a t G acts t r a n s i t i v e l y on ft a n d let X < ft. W e shall say that A is a J o r d a n set i f the pointwise stabilizer G is transitive on ft - X , and t h a t X is s t r o n g l y J o r d a n if, i n fact, G is p r i m i t i v e on ft-X. W e c a u t i o n the reader that this nomenclature is not standard; we have introduced it only for the m a t e r i a l leading up to the proof of T h e o r e m 8.18.
x x

It s h o u l d be obvious that if G acts on a set ft and X ft is an orbit for some subgroup H G, then for each element g G, the translate X - g is an orbit for the conjugate subgroup H . F u r t h e r m o r e , i f H is p r i m i t i v e on X , then H is p r i m i t i v e on X - g . A l s o , if X ft and g G G, then
9 9

244

8. P e r m u t a t i o n

Groups

( G ) = G -g. It follows from a l l of this that translates of J o r d a n sets are J o r d a n , a n d translates of strongly J o r d a n sets are strongly J o r d a n .
x x

8.21. L e m m a . Suppose t h a t a g r o u p G acts t r a n s i t i v e l y o n ft, a n d l e t X a n d Y be J o r d a n subsets such t h a t X U Y < ft. T h e n X n Y i s J o r d a n , a n d if b o t h X a n d Y a r e s t r o n g l y J o r d a n , t h e n X n Y i s s t r o n g l y J o r d a n . P r o o f . F i r s t , observe t h a t G and G are contained i n G nY, w h i c h acts on ft-(XnY). T h u s each G ^ - o r b i t and each G y - o r b i t on the set ft-(XnY) is contained i n a single G nY orbit. N o w ft - ( X n Y) = (ft - X ) U (ft - Y ) , a n d since X and Y are J o r d a n , it follows t h a t ft-X is a G - o r b i t and ft-Y is a G y - o r b i t . T o show t h a t G nY is transitive on ft - ( X n Y ) , therefore, it suffices to show t h a t ft-X a n d ft-Y are contained i n the same G y - o r b i t . T h i s follows, however, since (ft - X ) n (ft - Y) = ft - { X U Y) is nonempty. T h u s G x n Y is transitive on ft - ( X n Y ) , or i n other words, X n Y is a J o r d a n set, as desired.
x Y X X x X X n

Suppose now t h a t b o t h X and Y are strongly J o r d a n , and assume, as we may, that |X| < \Y\. T h e n |ft - X | > |ft - Y\, a n d since the sets ft - X a n d ft - Y are not disjoint, it follows that |ft - X | is more t h a n half the c a r d i n a l i t y of their u n i o n , w h i c h is ft - ( X n Y ) . N o w G stabilizes the set ft - X a n d acts p r i m i t i v e l y on it, a n d thus by L e m m a 8.20, the action of G x n Y on ft - ( X n Y ) (which we know to be transitive) is a c t u a l l y p r i m i t i v e . It follows t h a t X n Y is strongly J o r d a n , as required.
x

Suppose t h a t G is transitive on ft. If X ft is a subset obtained by deleting one point, then obviously X is a J o r d a n set. A l s o , the empty subset of ft is guaranteed to be J o r d a n . B u t if G is p r i m i t i v e on ft a n d there exists a nonempty J o r d a n set of c a r d i n a l i t y smaller t h a n |ft| - 1, we get a strong conclusion: the action of G on ft is d o u b l y transitive. 8.22. T h e o r e m . L e t G be a g r o u p a c t i n g p r i m i t i v e l y o n ft, a n d suppose that X C f t i s a J o r d a n set w i t h 0 < |X| < |ft| - 1. If a G ft i s a r b i t r a r y , t h e n G i s t r a n s i t i v e o n ft - { a } , a n d so G i s 2 - t r a n s i t i v e o n ft. F u r t h e r m o r e , if X i s s t r o n g l y J o r d a n , t h e n i n f a c t , G i s p r i m i t i v e o n ft - { a } .
a a

P r o o f . G i v e n the existence of the J o r d a n set X w i t h 0 < | X | < | f t | - l , we show t h a t every one-point subset of ft is J o r d a n . T h i s tells us t h a t for each point a G ft, the stabilizer G is transitive on ft - { a } , as wanted, and since G is transitive, L e m m a 8.2 guarantees t h a t G is d o u b l y transitive i n this situation. W e w i l l also show t h a t i f X is strongly J o r d a n , then every one-point subset of ft is strongly J o r d a n , a n d this w i l l complete the proof.
a

We can assume that X is m i n i m a l among nonempty J o r d a n subsets (or m i n i m a l among nonempty strongly J o r d a n subsets) and we work to show t h a t |X| = 1. Since translates of J o r d a n sets are J o r d a n , and translates of

8B

245

strongly J o r d a n sets are strongly J o r d a n , it w i l l follow t h a t every one-point subset of ft is J o r d a n (or is strongly Jordan) as wanted. F i r s t , assume t h a t |X| > |ft|/2, and let y = ft - X . N o w \Y\ > 1 since \X\ < |ft| - 1, and we have | Y | < |ft|/2 < |ft|. Since G is p r i m i t i v e on ft, it follows t h a t Y is not a block, and hence there exists an element g G G such t h a t Y-g ^ Y and Y n Y-g ^ 0 . Now X U X - g = (ft - Y) U (ft = ft - ( Y n y - y ) ,

and this is a proper subset of ft. Since X is J o r d a n , so too is X - g , and thus L e m m a 8.21 tells us t h a t X n X - # is J o r d a n . A l s o , i f X is strongly J o r d a n , t h e n by similar reasoning, X n X - is strongly J o r d a n .
5

N o w \X\ > |ft|/2 > \ X U A - | / 2 . T h e c a r d i n a l i t y of each of X a n d X-<y, therefore, exceeds half the c a r d i n a l i t y of their u n i o n , and it follows t h a t X H X - g is nonempty. Furthermore, X ^ X - g since y ^ Y - , and thus A n X - g < X . T h u s X n X - g is a n o n e m p t y J o r d a n (or strongly Jordan) set t h a t is s t r i c t l y smaller t h a n X , and this contradicts the m i n i m a l i t y of X .
5 5

We now have \X\ < |ft|/2, and assuming t h a t \X\ > 1, we work to derive a c o n t r a d i c t i o n . Since 1 < \X\ < |ft| a n d G is p r i m i t i v e , X cannot be a block, a n d thus there exists g G G such t h a t X ^ X - g a n d X n X - g ^ 0 . Since | X U X | < 2|X| < |ft|, we see t h a t X U X < ft, and so by L e m m a 8.21, the n o n e m p t y set X n X is J o r d a n (and it is strongly J o r d a n if X is). T h i s set is p r o p e r l y smaller t h a n X , however, a n d as before, this contradicts the m i n i m a l i t y of X .
9 9

F i n a l l y , we are ready to prove J o r d a n ' s theorem. P r o o f of T h e o r e m 8.18. W e can assume t h a t |ft| > 2 a n d t h a t | A | > 0, and we proceed by i n d u c t i o n on |ft|. B y hypothesis, the set A is strongly J o r d a n , and since | A | < |ft| - 2, we can a p p l y T h e o r e m 8.22 to deduce t h a t G is p r i m i t i v e on ft - { a } , where we take a G A .
a

N o w G is the pointwise stabilizer of the set A - { a } i n G , and by hypothesis, G is p r i m i t i v e on ft - A = (ft - a ) - (A - a ) . A l s o , | A - { a } \ < |ft - { a } | - 2, and so we can a p p l y the i n d u c t i v e hypothesis to the action of G to ft - { a } to deduce t h a t it is | A | - t r a n s i t i v e . B y L e m m a 8.2, therefore, G is ( | A | + Intransitive on ft, as wanted.
A a A a

A n o t h e r a p p l i c a t i o n of T h e o r e m 8.18 i n the spirit of T h e o r e m 8.17 a n d C o r o l l a r y 8.19 is the following. Its proof is somewhat more difficult, however. 8.23. T h e o r e m . L e t G be a p r i m i t i v e p e r m u t a t i o n g r o u p o n a set ft, a n d assume t h a t G c o n t a i n s a p - c y c l e , w h e r e p i s p r i m e a n d p < |ft| - 3. T h e n G i s e i t h e r t h e a l t e r n a t i n g g r o u p o r t h e f u l l s y m m e t r i c g r o u p o n ft.

246

8. P e r m u t a t i o n

Groups

We need an easy p r e l i m i n a r y result. 8.24. L e m m a . L e t x be a n n - c y c l e i n t h e s y m m e t r i c i s t h e f u l l c e n t r a l i z e r of x i n S .

n

group S.
n

Then (x)

P r o o f . A l l n-cycles i n S are conjugate, and it is easy to see that there are exactly (n - 1)! of t h e m . Since the size of the conjugacy class of x i n S is \S : C\, where C is the centralizer of x , we have (n - 1)! = n \ / \ C \ , a n d thus \ C \ = n . Since ( x ) C C and |(x>| = n , the result follows.
n n n

P r o o f of T h e o r e m 18.23. L e t A be the set of points fixed by the given p-cycle, a n d observe t h a t by C o r o l l a r y 8.19, we can assume t h a t p 3. N o w |ft - A | is the p r i m e number p , a n d since G contains a p-cycle, it follows t h a t G is transitive on ft - A , and i n fact it is p r i m i t i v e on this set. B y T h e o r e m 8.18, therefore, G is | A (-transitive on ft. ( A c t u a l l y , it is ( | A | + 1)transitive, but we do not need t h a t e x t r a information.) It follows that each p e r m u t a t i o n i n S y m ( A ) is induced by some element of G, or i n other words, G / G is n a t u r a l l y isomorphic to S y m ( A ) .
A A ( A ) A

N o w let P be the subgroup of order p generated by the given p-cycle. T h e n P C G , and since G acts faithfully on the p points of ft - A , we know t h a t | G | divides p \ . It follows t h a t P is a full Sylow p-subgroup of G , a n d since G < G , we can a p p l y the F r a t t i n i argument to deduce t h a t G = N G , where N = N ( P ) . W e conclude t h a t N / N = G /G = S y m ( A ) , and thus N induces the full s y m m e t r i c group on A .
A A A A A ( A ) ( A ) A G ( A ) A ( A ) A

Since | A | = |ft| - p > 3, it follows t h a t S y m ( A ) contains a 3-cycle, and it is easy to see t h a t such a 3-cycle must lie i n the derived subgroup of the s y m m e t r i c group. T h e r e exists, therefore, an element x e N ' such that the p e r m u t a t i o n induced by x on A is a 3-cycle. T h e a u t o m o r p h i s m group of P is abelian, and thus N / C is abelian, where C = C ( P ) . T h e n x e N ' C C , and so x commutes w i t h the given p-cycle. T h e p e r m u t a t i o n of ft - A induced by x , therefore, commutes w i t h a p-cycle on these p points. B y L e m m a 8.23, this p e r m u t a t i o n has order d i v i d i n g p , and hence x acts t r i v i a l l y o n ft - A . N o w x induces a 3-cycle on A , and since 3 does not divide p , it follows t h a t x also induces a 3-cycle on A . A t this point, we know t h a t x fixes the points of ft - A and it induces a 3-cycle on A . Since G acts faithfully on ft, we see t h a t x a c t u a l l y is a 3-cycle, a n d the result follows by C o r o l l a r y 8.19.
N p p p p

We close this section w i t h a result of A . B o c h e r t t h a t also relies on C o r o l l a r y 8.19. B o c h e r t ' s theorem shows t h a t a p r i m i t i v e p e r m u t a t i o n group t h a t is not either the alternating group or the full s y m m e t r i c group must actually be a fairly small subgroup of the s y m m e t r i c group. W e begin w i t h

8B

247

an elementary (but somewhat tedious) c o m p u t a t i o n t h a t enables us to find 3-cycles i n p e r m u t a t i o n groups. 8.25. L e m m a . L e t x a n d y be p e r m u t a t i o n s o n a set ft, a n d assume that t h e r e i s e x a c t l y o n e p o i n t a e ft m o v e d by b o t h x a n d y . T h e n t h e c o m m u tator [x,y] = x - ^ y - ^ x y is a 3-cycle. P r o o f . F i r s t , we observe that a , a-x a n d a-y are three distinct points. Certainly, a + a-x a n d a ^ a-y since x and y move a . A l s o , since a ^ a-x, we have a-y + { a - x ) - y = a-x, where the equality holds since a-x is different from a and is moved by x , and so it is fixed by y . N e x t , we show t h a t if 0 G ft is not one of the three points a , a-x or a-y, then p-ix'^y- ) = / ^ ( y ^ x " ) . T h e r e is n o t h i n g to prove if b o t h x ' and y " fix p, so by symmetry, it is no loss to assume t h a t x " moves P, and thus x moves p. Since x moves 0 a n d 0 + a , we know that y fixes 0, and so y " also fixes 0, and we have 0-{y- x~ ) = 0-x~ . A l s o , since x moves 0, it also moves 0-x~ , and since 0-x' + a , it follows that y fixes 0-x~ , and so y " also fixes this point. T h e n 0-( - y- ) = 0-x' , and this shows t h a t 0-(x- y~ ) = 0-{y- x~ ), as c l a i m e d . It follows t h a t
1 1 1 1 1 1 l l l l 1 l 1 1 1 1 X 1 1 l l

0 . [ x , y) = 0-{ - y~ xy)
1 l x

= 0-{y~ x' xy)

l 1

= 0 .

We have now shown that [ x , y ] fixes every member of ft other t h a n a , a-x and a-y. Now {a-x)-{x,y\ = a - { y - x y ) = a-{y- y) = a where the second equality holds because a-y' is moved by y and is different from a , and so it is fixed by x . It follows by interchanging the roles of x and y t h a t [ y , x ] carries a-y to a , and hence [ x , y ] = [ y , x ] ~ carries a to a-y. The points a-x, a and a-y are distinct, and we know t h a t [ x , y] carries the first of these to the second and the second to the t h i r d . Since [ x , y] fixes a l l other points i n ft, it must carry a-y back to a-x, a n d so [ x , y] is the 3-cycle (a-x, a , a-y).
l 1 l x

8.26. T h e o r e m ( B o c h e r t ) . L e t S = S y m ( f t ) , w h e r e |ft| = n , a n d suppose G i s a p r o p e r s u b g r o u p of S t h a t i s p r i m i t i v e o n ft. If G + A l t ( f t ) , t h e n \ S : G \ > [(n + l ) / 2 ] ! . P r o o f . Since the pointwise stabilizer S is t r i v i a l , there certainly exists a subset A C ft such t h a t S n G = 1. Choose A w i t h this property so t h a t TO = | A | as small as possible. N o w S acts faithfully on ft-A, and since S is the full s y m m e t r i c group on ft, it follows t h a t S induces the full s y m m e t r i c group on ft - A , and thus | 5 | = (n - TO)!.
a A A A A

248

8. P e r m u t a t i o n

Groups

We have \ S : G \ > \S
A

:GnS \
A

= \S \
A

= (n-m)!,

and so it suffices to show t h a t m < n/2, n - m > [(n+ l)/2].

since then n - TO > n / 2 , and thus

If TO > n / 2 , we w i l l show t h a t G contains a 3-cycle. T h e n since G is p r i m i t i v e , C o r o l l a r y 8.19 asserts t h a t G is either the alternating or symmetric group. T h i s is contrary to assumption, however, and t h a t w i l l complete the proof. A s s u m i n g t h a t m > n / 2 , we have |ft - A | = n - TO < TO = | A | , and so by the m i n i m a l i t y of | A | , it follows t h a t G n S n - A is n o n t r i v i a l . Choose a n o n i d e n t i t y element x G G n S n - A , and let a G ft be a point moved by x . T h e n a does not lie i n ft - A , a n d so a A . N o w | A - { a } | < | A | , and so by another appeal to the m i n i m a l i t y of | A | , we c a n choose a nonidentity element y G G n S _ . Since G n 5 = 1, it follows t h a t y does not lie i n S , and thus y must move a . T h u s x and y b o t h move a , but there can be no other point moved b y b o t h x and y since if 0 7^ a , then either 0 G ft - A , and so 0 is fixed by x , or else 0 G A - { a } a n d /? is fixed by y . B y L e m m a 8.25, the commutator [ x , y ] is a 3-cycle, and the proof is complete.
A { a } A A

Problems

8B

8 B . 1 . L e t G be transitive on ft, and let a G ft. F i x a nonempty subset X C ft and let A be the intersection of a l l of those translates of X t h a t contain a . Show t h a t A is a block. 8 B . 2 . L e t G be p r i m i t i v e on ft, and let a,0 G ft w i t h a ^ 0. If X C ft is n o n e m p t y a n d proper, show that there exists an element g G G such that a - g G X a n d 0-g 0 X . 8 B . 3 . L e t G be a p r i m i t i v e p e r m u t a t i o n group on ft, and suppose t h a t M and N are distinct m i n i m a l n o r m a l subgroups of G. P r o v e t h a t the following hold. (a) M and N are regular subgroups. (b) M = C
G

( N ) and N =

C { M ) .
G

(c) M = N . H i n t . Use P r o b l e m 8 A . 4 . 8 B . 4 . L e t G be a solvable p r i m i t i v e p e r m u t a t i o n group. Show t h a t the degree of G is a p r i m e power and t h a t G has a unique m i n i m a l n o r m a l subgroup.

Problems 8 B

249

8 B . 5 . L e t G be a p r i m i t i v e p e r m u t a t i o n group on ft, and let H = G be a point stabilizer. If H has a fixed point i n ft - { a } , show t h a t \ H \ = 1, and t h a t G has p r i m e order.
Q

8 B . 6 . L e t G be a p r i m i t i v e p e r m u t a t i o n group on ft, and let H = G be a point stabilizer. If # has a n orbit of size 2 on ft - { a } , show t h a t \ H \ = 2 a n d t h a t G = /J> , the dihedral group of order 2 p , where p > 2 is a p r i m e number.
a 2p

8 B . 7 . L e t G be a p r i m i t i v e p e r m u t a t i o n group on ft, a n d let H = G be a point stabilizer. If H has an orbit of size 3 on ft - { a } , show t h a t \ H \ has the form 3-2 for some integer e > 0.
a e

H i n t . L e t 5 lie i n a n P - o r b i t of size 3 and let D = H p . Show t h a t 0 ( > ) < 0 { K ) , where K = c o v e { D ) .

2 H

N o t e . B y a result of C . Sims, e < 4, a n d so \ H \ divides 48. Sims conjectured t h a t if a point stabilizer i n a p r i m i t i v e p e r m u t a t i o n group has an orbit of size r , t h e n the order of the point stabilizer is b o u n d e d by some function of r . Sims' conjecture was eventually proved by C a m e r o n , Praeger, S a x l a n d Seitz, but unfortunately their proof (which is the o n l y one known) relies o n the classification of simple groups. 8B.8. Let x carries i to i suppose t h a t and fixes a l l primitive. S be the n-cycle ( l , 2 , . . . , n - l , n ) . (In other words, x + 1 for 1 < i < n , and it carries n to 1.) L e t 1 < m < n a n d H C S is a subgroup t h a t acts t r a n s i t i v e l y o n { 1 , 2 , . . . , m } of the r e m a i n i n g points. Show t h a t the group G = ( H , x ) is
n n

H i n t . G i v e n a block A w i t h | A | > 1, show t h a t some translate of A contains a point i < m and also a point j > m. Deduce t h a t t h a t translate contains all i w i t h 1 < i < m. 8 B . 9 . L e t x G S be the n-cycle ( l , 2 , . . . , n ) , and let y be the m-cycle ( 1 , 2 , . . . , m ) , where 1 < m < n . Show t h a t G = ( x , y ) is the full s y m m e t r i c group S if either of the integers m and n is even, and otherwise, G is the alternating group.
n n

8 B . 1 0 . Suppose that G C S stabilizer. Show t h a n n = 6.

is a subgroup of index n t h a t is not a p o i n t

H i n t . Show t h a t G is transitive, a n d then use P r o b l e m 8 A . 1 6 to deduce t h a t it is 2-transitive, and thus p r i m i t i v e . A p p e a l to B o c h e r t ' s theorem to deduce t h a t n > [(n + l ) / 2 } \ . Show t h a t the o n l y positive integers n for w h i c h this inequality is v a l i d are 1, 2, 3, 4 and 6.

250

8. P e r m u t a t i o n

Groups

N o t e . T h e s y m m e t r i c group 5 a c t u a l l y does have a subgroup of index 6 t h a t does not stabilize a point. T o see this, observe t h a t S acts transitively on its six Sylow 5-subgroups. It is not h a r d to show t h a t this is a faithful action, and thus S can be embedded as a transitive subgroup of 5 , and of course, this copy of S i n 5 has index 6 and does not stabilize a point.
6 5 5 6 5 6

N o t e also t h a t i f S has a non-inner a u t o m o r p h i s m a , then a w o u l d carry a point stabilizer to a subgroup of index n t h a t is not a point stabilizer, and by this p r o b l e m , t h a t cannot h a p p e n unless n = 6. It follows t h a t a l l automorphisms of S are inner for n 6. Furthermore, it is easy to show t h a t a subgroup H of index 6 i n 5 t h a t does not stabilize a point has t r i v i a l core, and this there is an injective h o m o m o r p h i s m from S to S that maps i f to a point stabilizer. T h i s m a p is a non-inner a u t o m o r p h i s m of S .
n n 6 6 6 6

8C T h r o u g h o u t m u c h of this book we have sought results t h a t could be used to prove t h a t a group is not simple, but i n this section, we w i l l use some permutation-group theory to prove that certain groups actually are simple. We begin w i t h a comparatively easy result of this type. 8 . 2 7 . T h e o r e m . If n > 5, t h e a l t e r n a t i n g g r o u p A
n n

is simple.
n

P r o o f . A s s u m i n g t h a t 1 < N < A , we work to show t h a t N = A . Since A is (n - 2)-transitive i n its n a t u r a l action, it is certainly 2-transitive, and thus it is p r i m i t i v e by L e m m a 8.16. B y L e m m a 8.15, therefore, N acts transitively.
n

Suppose first that n = 5. Since N acts transitively on a set of c a r d i n a l i t y 5, we deduce t h a t 5 divides \ N \ . B u t \ A \ = 60, and so \ A : N \ is not divisible by 5, and thus a l l 5-cycles i n A lie i n N . T h e number of these 5-cycles is 24, a n d thus \ N \ > 24. Since \ N \ divides 60, the o n l y possibilities are \ N \ = 30 or \ N \ = 60, and, i n particular, \ A : N \ is not divisible by 3. T h e n a l l 3-cycles i n A lie i n N , and we count t h a t there are 20 of these. Since N contains a l l 5-cycles and a l l 3-cycles, we have \ N \ > 24 + 20 = 44. The only possibility is t h a t \ N \ = 60, and thus N = A and A is simple.
5 5 5 b b 5 b

We can now suppose t h a t n > 6, and we proceed by i n d u c t i o n on n to derive a c o n t r a d i c t i o n i f N < A . L e t H be a point stabilizer i n A , and observe t h a t H = A a n d so H is simple by the inductive hypothesis. A l s o , since N is transitive, we have N H = A by L e m m a 8.5. T h e n H % N since otherwise, we w o u l d have N = A , w h i c h we are assuming is not the case. It follows t h a t N n H is a proper n o r m a l subgroup of H , and since H is simple, we have N n H = 1. Since N is transitive, it follows t h a t it is a regular n o r m a l subgroup of A , and by L e m m a 8.5, the conjugation
n n n U n n n

8C

251

action of H on the nonidentity elements of TV is permutation-isomorphic to the n a t u r a l action of H = A - on n - 1 points. T h e n | i V | = n , and the action of H on its nonidentity elements is (n - 3)-transitive, and, i n particular, it is 3-transitive. However, this is impossible since by L e m m a 8.8(c), the 3-transitive action of H on the n o n i d e n t i t y elements of TV w o u l d force |TV| = 4 .
n X

8 . 2 8 . C o r o l l a r y . If n > 5, t h e s y m m e t r i c g r o u p S subgroups: the identity, the whole group a n dA

n n

has e x a c t l y t h r e e n o r m a l

.
n n n

P r o o f . Suppose TV < S and that TV is not one of A or S . T h e n TV 2 A , and thus TV n A is a proper n o r m a l subgroup of the simple group A . It follows t h a t J V n A = l , a n d thus \ N \ < \S : A \ = 2. Since S is n transitive on n-points, it is p r i m i t i v e , a n d thus if TV > 1, t h e n TV is transitive a n d |TV | > n . T h i s is not the case, however, a n d so TV = 1.
n n n n n n

We t u r n now to the projective special linear groups P S L ( n , q ) . R e c a l l t h a t by definition, P S L ( n , q ) = S L { n , q ) / Z , where Z is the central subgroup of S L ( n , q ) consisting of the scalar matrices w i t h determinant equal to 1. We begin work now t o w a r d a proof t h a t w i t h just two exceptions, a l l of the groups P S L ( n , q ) are simple for p r i m e powers q a n d integers n > 2. T h e exceptions are P S L ( 2 , 2 ) and P S L ( 2 , 3 ) , w h i c h are not simple. (In fact, these two groups are solvable since P S L ( 2 , 2 ) has order 6 a n d is isomorphic to the s y m m e t r i c group S , and PSL(2,3) has order 12 a n d is isomorphic to the alternating group A . ) A c t u a l l y , the groups P S L ( n , F ) are also simple when F is an infinite field. A l t h o u g h the proof t h a t we w i l l present for finite fields goes t h r o u g h w i t h o u t essential m o d i f i c a t i o n for infinite fields too, we w i l l nevertheless l i m i t our discussion to the finite case.
3 4

A s s u m i n g t h a t n > 2 (as we w i l l , throughout this discussion), it follows by L e m m a 8.3 t h a t S L { n , q ) acts 2-transitively on the set ft of onedimensional subspaces of a vector space V of dimension n over the field F of order q. Since the action of a scalar m a t r i x on the vector space V is just scalar m u l t i p l i c a t i o n , the subgroup Z acts t r i v i a l l y on ft. T h u s Z is contained i n the kernel of the action, a n d we can view P S L ( n , q ) = S L ( n , q ) / Z as acting on ft. I n fact, this action is faithful. 8 . 2 9 . L e m m a . F o r a l l i n t e g e r s n > 2 a n d a l l p r i m e p o w e r s q, t h e g r o u p PSL{n,q) i s a 2 - t r a n s i t i v e p e r m u t a t i o n g r o u p of d e g r e e ( q - l ) / ( g - 1).
n

P r o o f . W e k n o w that P S L ( 2 , q ) acts on the set ft of one-dimensional subspaces of the n - d i m e n s i o n a l space V over the field F of order q. N o w V contains exactly q - 1 nonzero vectors, a n d each of these lies i n a unique one-dimensional subspace. Since each such subspace contains exactly q - 1 nonzero vectors, it follows that |ft| - ( q - l ) / ( g - 1).
n n

252

8. P e r m u t a t i o n

Groups

It remains to show t h a t P S L ( n , q ) acts faithfully on ft, or equivalently, t h a t Z is the full kernel of the action of S L { n , q ) on ft. T o see this, let x G S L ( n , q ) lie i n the kernel of the action. T h e n x fixes every one-dimensional subspace of V, a n d thus x carries every vector v G V to some scalar m u l t i p l e of itself. Let { v i | 1 < i < n } be a basis for V, and w r i t e ( )x = am for appropriate scalars o n G F . T h e n x corresponds to a diagonal m a t r i x , and to complete the proof, it suffices to show t h a t a = a , for 1 < i < j < n . N o w for some scalar a G F , we have
Vi {

a ( v i + Vj) = { v i + V j ) x = ViX + v j x = am a n d since v required.

{

+ ajVj

,
a j

and

V j

a r e linearly independent, this yields o n = a =

, as

R e c a l l t h a t a group G is said to be perfect i f G' = G. A l s o , i f S C G is an a r b i t r a r y subgroup, then the n o r m a l closure of S i n G is the subgroup S = (S \ g <E G ) ; it is the unique smallest n o r m a l subgroup of G t h a t contains S. O f course, i f G is a nonabelian simple group, then G is perfect and S = G for every n o n t r i v i a l subgroup S. Conversely, if G is perfect a n d it is the n o r m a l closure of an appropriate subgroup, a n d if, i n a d d i t i o n , G is a p r i m i t i v e p e r m u t a t i o n group, our next result guarantees t h a t G is simple. Since P S L ( n , q ) is a d o u b l y transitive p e r m u t a t i o n group by L e m m a 8.29, it is certainly a p r i m i t i v e p e r m u t a t i o n group, and hence the following l e m m a of K . Iwasawa is relevant to establishing the s i m p l i c i t y of this group.
G 9 G

8 . 3 0 . L e m m a (Iwasawa). L e t G be a p r i m i t i v e p e r m u t a t i o n g r o u p , a n d assume t h a t G' = G. L e t H = G be a p o i n t s t a b i l i z e r i n G, a n d suppose t h a t t h e r e exists a s o l v a b l e s u b g r o u p A < H such that G = A . Then G is simple.
a G

P r o o f . L e t 1 < N < G, and observe t h a t since G is a p r i m i t i v e p e r m u t a t i o n group, N is transitive, a n d thus N H = G. N o w N < G , a n d since A < H , we see t h a t H C N ( N A ) . A l s o N C N ( N A ) , and thus since N H - G, we have N A < G. N o w A = G and A C N A < G, and it follows that N A = G. B u t A is solvable, so G / N must also be solvable, and hence i f N < G, we have ( G / N ) ' < G/N. Since G' = G, however, we see t h a t ( G / N ) ' = G/N, a n d this c o n t r a d i c t i o n shows t h a t N = G. T h i s completes the proof.
G G G

W e w i l l use L e m m a 8.30 to prove t h a t a group of the form G = P S L ( n , q ) is simple (except when n = 2 and q < 3). T o do this, we w i l l establish (in the non-exceptional cases) t h a t G' = G, and we w i l l find a solvable (and, i n fact, a n abelian) n o r m a l subgroup A of a point stabilizer, such that A = G. (Note t h a t the groups P S L ( 2 , 2 ) and P S L ( 2 , 3 ) definitely are not perfect

8C

253

since they are solvable.) T h e key to b o t h of these tasks is to consider certain special matrices t ^ { a ) i n S L ( n , q ) . T o define the matrices t ^ { a ) a n d to facilitate m a t r i x computations, we recall t h a t the m a t r i x unit' e is the m a t r i x whose only nonzero entry is 1, o c c u r r i n g m the (t,j)-positi'on. It is t r i v i a l to check t h a t m a t r i x units m u l t i p l y according to the formula
i,v

i f J'

otherwise.

It is the s i m p l i c i t y of this rule t h a t makes the m a t r i x units so useful for computation. G i v e n 1 < i, j < n w i t h i + j , define U j ( a ) = 1 + a e i j , where, i n this formula, 1 is the n x n identity m a t r i x and a G F is nonzero. Clearly, t { a ) is either upper or lower triangular (depending on whether i < j or i > j ) and its diagonal entries are a l l 1. T h e determinant of this m a t r i x , therefore, is 1, and so U A a ) G S L { n , q ) , as wanted.
i t j

We digress briefly to put our matrices t ( a ) into a somewhat more general context. L e t V be a vector space of d i m e n s i o n n > 2. A h y p e r p l a n e i n V is a subspace of dimension n - 1, or equivalently, a subspace of codimension 1. A linear operator t on V is a t r a n s v e c t i o n i f its fixed-point space is a hyperplane W, and t acts as the i d e n t i t y on V/W. It is not h a r d to show t h a t a l l of our matrices U j ( a ) are transvections and also that the transvections i n G L ( n , q ) form a single conjugacy class. I n p a r t i c u l a r , a l l of the transvections i n G L ( n , q ) a c t u a l l y lie i n S L ( n , q ) , a n d one can show t h a t if n > 3, a l l of these transvections are conjugate i n S L ( n , q ) .
i d

We r e t u r n now to work t o w a r d our m a i n goal, w h i c h is the proof t h a t the groups P S L { n , q ) are simple (except w h e n they are not). 8.31. T h e o r e m . L e t n > 2. T h e n t h e g r o u p S L ( n , q ) m a t r i c e s t ( a ) w i t h i ^ j a n d a ^ 0.
i d

i s g e n e r a t e d by t h e

P r o o f . If a is any n x n m a t r i x over the field F of order g, it is easy to see that the p r o d u c t m a t r i x t { a ) a can be obtained from a b y adding a times row j to row i, and, similarly, the m a t r i x a t ^ { a ) c a n be obtained from a by a d d i n g a times c o l u m n i to c o l u m n j . W e describe an a l g o r i t h m by w h i c h an a r b i t r a r y m a t r i x a e S L ( n , q ) c a n be converted into the identity m a t r i x by a sequence of row and c o l u m n operations of these types, and it follows t h a t 1 = p a q , where each of p and q is a p r o d u c t of matrices of the form i i j ( a ) . T h i s w i l l prove the result.
i d

Let a G S L { n , q ) . T h e first c o l u m n of a is not zero, and since n > 2, some row operation of the above type (if necessary) w i l l convert a to a m a t r i x i n w h i c h the (z, l ) - e n t r y is a nonzero field element 5, where i > 1. A s s u m i n g

254

8. P e r m u t a t i o n

Groups

t h a t has been done, let 7 be the (1, l ) - e n t r y of the resulting m a t r i x , where possibly, 7 = 0. If we add { l - - y ) / S times row i to row 1, we o b t a i n a m a t r i x w i t h (1, l ) - e n t r y equal to 1. N o w a sequence of additions of appropriate multiples of the first row to other rows w i l l y i e l d a m a t r i x w i t h (1, l ) - e n t r y equal to 1 a n d a l l other entries i n the first c o l u m n equal to 0, and then a sequence of additions of appropriate multiples of the first c o l u m n to other columns w i l l y i e l d a m a t r i x where the (1, l ) - e n t r y is 1 and a l l other entries i n the first row a n d first c o l u m n are 0. Observe that the m a t r i x constructed i n the previous paragraph has determinant 1 since a and a l l of the matrices t { a ) have determinant 1. W e change n o t a t i o n now, a n d call this new m a t r i x a . If n = 2, then a is diagonal, a n d since its (1,1) entry is 1 a n d its determinant is 1, the (2,2)-entry of a must also be 1, a n d thus a is the identity m a t r i x , as wanted. A s s u m e then, t h a t n > 3. Since some entry i n c o l u m n 2 is nonzero, we can do row and c o l u m n operations as before, but using only rows a n d columns numbered 2 or larger. T h i s w i l l not change either the first row or the first c o l u m n of a , but we can arrange t h a t the (2,2)-entry of the resulting m a t r i x is 1, and t h a t a l l other entries i n the second row a n d second c o l u m n are 0.
i d

If n = 3, the resulting m a t r i x is diagonal, and so It is the identity m a t r i x by the determinant argument t h a t we used before. If n > 3, we again do row and c o l u m n operations, but this t i m e w i t h rows and columns numbered 3 or higher. C o n t i n u i n g like this, we eventually o b t a i n the identity m a t r i x , and this completes the proof. 8 . 3 2 . T h e o r e m . / / n > 3 o r n = 2 a n d q > 3, t h e n S L { n , q ) i s perfect.
i d

P r o o f . B y T h e o r e m 8.31, it suffices to show t h a t the m a t r i x t ( a ) is a c o m m u t a t o r , where i + j and a ^ 0. W e begin w i t h the observation that if i j , then ( e ^ ) = 0. T h e n (1 - a ) ( l + a e ) = 1, a n d hence (l + a e i j ) - = l - a .
2 e i j 1 e i J

Suppose first t h a t n > 3, a n d let 1 < i , j , k < n , where i, j and k are distinct. T h e n (ei,fc) a n d thus (1 + e It follows t h a t [(1 + e
i>fc i j k 1+aCfc

'

= (1 - a e

k t j

){e

i i k

) ( l + ae )
kJ

= e

i)fc

+ ae

hJ

1 + a e

= 1+ e

ijk

+ ae

id

), (1 + a e

k J

) ] = (1 + e = (l-ei
i

y {l +
k

) ( l + ei, + a e i j )

= 1+ aeij
= t i , j ( )>
a

8C

255

and thus t We

i t j

( a ) is a commutator, as wanted.

can now assume that n = 2, and we let (3,7 e F be nonzero. L e t 6 and Then c =

so t h a t b and c lie i n S L ( 2 , q ) . M

"/T
.
0

p .

0"
0

"1 0

7 1

_ 1

7T 0

0" "1 iij 0

7 1

1 0

-7 1

Q
j

P - P i 0 /r .
1

/5- 7

. 0
2

1 0
2

7(1-/3 ) 1
2

Now let a be an a r b i t r a r y nonzero element of F . If q > 3, choose (3 / 1 . T h e n 1 - /3 ^ 0, and so we can choose 7 so t h a t 7(1 - ( 3 ) = a . T h e n [ b , ] = i i , ( a ) , a n d so i i , ( a ) is a c o m m u t a t o r i n 5 L ( 2 , g ) , as wanted. B y t a k i n g transposes, we see t h a t i ( a ) is also a c o m m u t a t o r i n S L { 2 , q ) , and this completes the proof.
c 2 2 2 j l

We 8.33.

can now establish the m a i n result of this section. T h e o r e m . / / n > 3 o r n = 2 a n d q > 3, t h e n P S L ( n , q ) is simple.

P r o o f . U s i n g the bar convention, we can write P S L ( n , q ) = S = S/Z, where S = S L ( n , q ) and Z is the group of scalar matrices i n S. L e t V be the n-dimensional row space over F , a n d let { v \ 1 < i < n } be the standard basis for V, so that v is the row vector ( 0 , . . . , 0 , 1 , 0 , . . . , 0), whose only nonzero entry is i n p o s i t i o n i. N o w S acts by right m u l t i p l i c a t i o n on V, and the corresponding action of S on the set ft of one-dimensional subspaces of V is faithful a n d d o u b l y transitive by L e m m a 8.29. In particular, S is a p r i m i t i v e p e r m u t a t i o n group on ft. A l s o , since S is perfect by T h e o r e m 8.32, we have ( S ) ' = S = S, and so S is perfect.
{ t J

Now fix an integer j w i t h 1 < j < n , and let H be the stabilizer i n S L ( n , q ) of the one-dimensional space Uj = F . T h e n H j acts on the vector space V / U j , and this induces a group h o m o m o r p h i s m from H j into the group G L ( V / U j ) of invertible linear operators on V/U L e t A j be the kernel of this h o m o m o r p h i s m , so t h a t A j is the group of a l l matrices in S L ( n , q ) t h a t stabilize Uj and act t r i v i a l l y on V / U j . In particular, if a A j , then det(a) = 1, a n d since the linear transformation induced by a on V / U j is the identity m a p , it too has determinant 1. It follows t h a t the
3 V j r

256

8. P e r m u t a t i o n

Groups

linear transformation of Uj induced by a also has determinant 1. T h e m a p induced by a on the one-dimensional space Uj is thus the i d e n t i t y map, and so ( j ) a = Vj. A l s o , since a fixes every element of V / U j , we see t h a t if i ^ j then ( ) a = v + favj for some scalar fa depending on i . I n other words,
V V i {

a =

1 +

fc id
e

'

and it is easy to see t h a t A j is exactly the set of matrices of this form. It follows by the m u l t i p l i c a t i o n rule for m a t r i x units t h a t A is a n abelian group, and, of course, A j < H j since A j is the kernel of a h o m o m o r p h i s m defined on H j . N o t i c e also that the matrices t { a ) lie i n A j , for i ^ j .
3 i t j

Now S = S L ( n , q ) acts t r a n s i t i v e l y on the set of one-dimensional subspaces of V, and since the group A j is uniquely determined by the space Uj = F v j , it follows that a l l of the abelian groups A j are conjugate i n S. (We are not saying, however, that they form a full conjugacy class of subgroups.) E v e r y m a t r i x of the form ( ) lies i n one of the conjugate subgroups A j , and since these matrices generate S by T h e o r e m 8.31, it follows t h a t ( A ) = S.
5 x

Now

H [ is a point stabilizer i n the p r i m i t i v e group S, and I T is a solvable (It is irrelevant t h a t A

x

(and i n fact abelian) n o r m a l subgroup o f f f j . does not c o n t a i n Z . ) A l s o , ( A [ )

s

usually

= {Ai)

= S. Since S is perfect, it follows

by L e m m a 8.30 t h a t it is simple.

We m e n t i o n t h a t the s i m p l i c i t y of P S L ( n , q ) implies t h a t the scalar subgroup Z C S L ( n , q ) is a c t u a l l y the full center of S L ( n , q ) . Problems 8C 8C.1. Show t h a t A has no subgroup of order 120, and deduce t h a t a group of order 120 cannot be simple.
6

N o t e . T h e r e does exist a perfect group of order 120, however: S L ( 2 , 5). 8C.2. Suppose t h a t G is a p e r m u t a t i o n group of degree 11 a n d order 7920 =

1 M 0 - 9 - 8 . P r o v e t h a t G is simple. H i n t . Show t h a t | N ( P ) | = 55, where P G S y L l ( G ) .

G

N o t e . I n fact, G must be isomorphic to the M a t h i e u group M 8C.3.

Suppose t h a t G is a transitive p e r m u t a t i o n group of degree 12 and

order 95,040 = 1 2 - 1 M 0 - 9 - 8 . P r o v e that G is simple. H i n t . Show t h a t G must be 2-transitive, and t h a t a nonidentity proper n o r m a l subgroup w o u l d have to be regular. Observe t h a t a group of order 12 can never be a m i n i m a l n o r m a l subgroup of any group.

8D

257

N o t e . In fact, G must be isomorphic to the M a t h i e u group

1 2

8C.4. L e t G be a transitive group of degree 100 on a set ft, and suppose t h a t a point stabilizer G is simple, a n d t h a t it has orbits of size 22 a n d 77 on ft - {a}. Show t h a t G is p r i m i t i v e , a n d deduce t h a t it is simple.
a

N o t e . T h e H i g m a n - S i m s group H S satisfies these conditions, and, i n that case, the point stabilizer H is isomorphic to the M a t h i e u group M .
2 2

8 C . 5. L e t G be a 5-transitive group of degree 24, a n d assume that the setwise stabilizer of three points is simple. P r o v e t h a t the stabilizer i n G of two points, the stabilizer i n G of one point a n d G itself are simple groups. N o t e . T h i s p r o b l e m describes the M a t h i e u groups M , M and M . The three-point stabilizer is isomorphic to P S L ( 3 , 4 ) , acting 2-transitively on 21 = ( 4 - l ) / ( 4 - 1) points, as i n L e m m a 8.29.
2 2 2 3 2 4 3

8C.6. L e t A be an abelian group. Show t h a t A u t ( A ) is simple i f and only if A has order 3 or A is an elementary abelian 2-group of order at least 8. 8D We assume throughout this section t h a t G acts t r a n s i t i v e l y on ft, a n d we study the orbits of the action of a point stabilizer on ft. A s we shall see, a good way to understand these s u b o r b i t s is to consider the n a t u r a l componentwise action of G on the set ft x ft of ordered pairs of points, where this action is denned by the formula ( a , 0 ) - g = (a-g,0-g). T h e set ft x ft is partitioned into G - o r b i t s , w h i c h are usually called orbitals. One of these is the diagonal {(a, a) | a G ft}. (The diagonal is certainly G-invariant, a n d it is a single G - o r b i t because G is transitive on ft.) If |ft| > 1, the diagonal obviously does not exhaust ft x ft, a n d so the t o t a l number of G-orbits on this set (in other words, the t o t a l n u m b e r of orbitals) is at least 2. T h e number of orbitals is the r a n k of the action, w h i c h is defined only for transitive actions. If A is an o r b i t a l , it is obvious t h a t the set A ' = {(/?, a) I ( a , 0 ) G A} is also an o r b i t a l , a n d we say t h a t A a n d A ' are p a i r e d orbitals. O f course, an o r b i t a l may be paired w i t h itself. F o r example, the diagonal o r b i t a l is self-paired, and i f the rank is 2, there is only one other o r b i t a l , so it too must be self-paired. A transitive action of G on ft has rank 2 precisely w h e n |ft| > 1 a n d G acts transitively on the set of a l l of pairs i n ft x ft t h a t lie outside of the diagonal. These, of course, are exactly the pairs of d i s t i n c t points i n ft, a n d we recall that by definition, G is d o u b l y transitive on ft i f and only i f it

258

8. P e r m u t a t i o n

Groups

acts transitively on this set of pairs. I n other words, a transitive action is 2-transitive i f a n d only if it has rank 2. If point More equal G has rank 2 on ft, it is 2-transitive, and, i n this case, we know that a stabilizer G has exactly two orbits on ft, namely, { a } a n d ft - {a}. generally, we shall see that the number of G - o r b i t s on ft is always to the rank.
a a a

It should not be surprising t h a t the number of orbits of G on ft is independent of the choice of a since by transitivity, the various point stabilizers are conjugate i n G. W e might even expect t h a t the sizes of these suborbits correspond as a varies. I n fact, even more is true. G i v e n points a , 0 G ft, there is a n a t u r a l , size-preserving bijection between the sets of G - o r b i t s and G/3-orbits on ft. T o define this bijection, we introduce some notation.
Q

Let

A be an o r b i t a l . W e write A(a) =

{0

ft |

(a,0)

G A},

so t h a t the o r b i t a l A defines a function from ft into the set of a l l subsets of ft. These o r b i t a l f u n c t i o n s are i n t i m a t e l y connected w i t h the suborbits. 8 . 3 4 . L e m m a . L e t G be a g r o u p a c t i n g t r a n s i t i v e l y o n ft, a n d l e t a i n ft. T h e n as A r u n s o v e r t h e o r b i t a l s , t h e sets A (a) a r e d i s j o i n t , a n d they are e x a c t l y t h e o r b i t s of G o n ft. A l s o , if g G G, t h e n f o r each o r b i t a l A , w e h a v e A ( a - g ) = A ( a ) - g . I n p a r t i c u l a r | A ( a ) | i s i n d e p e n d e n t of a G ft, a n d , i n fact, \ A { a ) \ = |A|/|ft|.
a

P r o o f . Suppose t h a t A ( a ) and A ( a ) have some point 0 m c o m m o n , where A and A are orbitals. T h i s means t h a t the pair { a , 0 ) lies i n each of the sets A and A , and since these sets are G-orbits on ft x ft and they are not disjoint, they must be equal. It follows t h a t the sets A ( a ) are disjoint for distinct orbitals A .
x 2 x 2 x 2

Now let g G G, and suppose A is an o r b i t a l . Since A is a G-invariant subset of ft x ft, it follows t h a t { a , 0) G A if and only if { a - g , 0-g) G A . E q u i v a l e n t s , 0 A ( a ) i f and only if 0-g G A ( a - g ) , a n d this proves t h a t A ( a - g ) = A { a ) - g , as wanted. A l s o , since a - g runs over a l l of ft as g varies, it follows t h a t for a fixed o r b i t a l A , a l l of the sets A ( a ) have equal cardinality. If g G G , we have A { a ) - g = A ( a - g ) = A ( a ) , and thus A ( a ) is a G invariant subset of ft. T o show t h a t it is actually a G - o r b i t , let 0,7 G A { a ) . T h e n ( a , 0 ) a n d (a, 7) lie i n A , w h i c h is a G - o r b i t on ft x ft. There exists, therefore, a n element g G G t h a t carries ( a , 0) to (a, 7). T h e n 0-g = 7, and since a - g = a , w e have g G G , as wanted.
a a Q a

8D

259

F i n a l l y , for a ft, we see that | A ( a ) | is exactly the number of pairs i n A whose first entry is a , and thus |A| = ^|A(a)|.
a&fl

B u t | A (a) | is independent of a , and this yields | A | = | f t | | A ( a ) | .

A s we indicated, L e m m a 8.34 establishes a n a t u r a l bijection from the set of G - o r b i t s o n ft onto the set of G ^ - o r b i t s on ft, where a , 0 ft. G i v e n a G - o r b i t X , we k n o w t h a t X = A ( a ) for some unique o r b i t a l A , a n d we let X correspond to the G ^ - o r b i t Y = A (/J). In fact, given X , it is easy to determine Y w i t h o u t k n o w i n g the o r b i t a l A . T h i s is because 0 = a - g for some element g G , a n d so
Q a

Y = A ( 0 ) = A(a-g) = A(a)-g = X - g. T h e cardinalities of the orbits of G for any given point a are called the s u b d e g r e e s of G , and this list of positive integers is independent of the choice of a . A l s o , at least one of the subdegrees is the number 1 since { a } is an orbit of G . (Note that this t r i v i a l G - o r b i t is A ( a ) , where A is the diagonal orbital.)
a a a

In fact, the corresponding orbits X a n d Y for G a n d G share more t h a n cardinalities. F i x g G c a r r y i n g a to 0. T h e n the i s o m o r p h i s m u ^ v P from G to G and the bijection x ^ x - g from X to Y are compatible i n the sense t h a t
a 0 a 0

{x-u)-g =

(x-g)-u .

It is easy to see from this t h a t a property such as p r i m i t i v i t y , 2-transitivity, faithfulness, etc. holds for the action of G on X i f and o n l y if it holds for the a c t i o n of G p on Y. (We hope t h a t the reader w i l l forgive us for o m i t t i n g a formal proof of this.)
Q

A useful way to t h i n k about orbitals is v i a g r a p h theory. W e can view a n o r b i t a l A as the set of edges of a directed g r a p h w i t h vertex set ft. T o do this, we imagine ft to be set of points i n space, a n d we draw an arrow from a t o 0 precisely when the pair ( a , 0 ) lies i n A . ( T h i s is not very interesting if A is the diagonal, but i n a l l other cases, each arrow joins a pair of distinct points.) T h e graph constructed i n this way is the o r b i t a l g r a p h associated w i t h A , a n d we w i l l refer to its edges A - a r r o w s . Since every ordered pair of points i n ft lies i n exactly one o r b i t a l , it follows t h a t for each such pair ( a , 0 ) , there is a unique o r b i t a l A such t h a t A - a r r o w goes from a to 0. Perhaps a g o o d way to t h i n k about this is to imagine the orbitals as colors. There is a n arrow j o i n i n g a t o 0 for every choice of a , 0 e ft, a n d the "color" of t h a t arrow is determined by the o r b i t a l

260

8. P e r m u t a t i o n

Groups

t h a t contains the corresponding pair ( a , 0 ) . ( T h i s point of view is convenient w h e n we need to t h i n k about two or more o r b i t a l graphs simultaneously.) Observe that for each o r b i t a l A , the set of points reached by A - a r r o w s leaving a is exactly A (a). A l s o , a n d most i m p o r t a n t l y , i f there is a A - a r r o w from a to 0, then there is also a A - a r r o w from a - g to 0-g. (This, of course, is because A is a G - i n v a r i a n t subset of ft x ft.) It follows that the given action of G on ft defines automorphisms of each of the o r b i t a l graphs. O f course, G acts t r a n s i t i v e l y on the points i n the A - g r a p h , but it also is edge-transitive because A is a G - o r b i t on ft x ft. The following result provides a pretty connection between the graph theory a n d the group theory. 8.35. T h e o r e m . L e t G be a g r o u p a c t i n g t r a n s i t i v e l y o n ft. T h e n G i s p r i m i t i v e if a n d o n l y if t h e g r a p h a s s o c i a t e d w i t h every n o n - d i a g o n a l o r b i t a l is connected. The statement of T h e o r e m 8.35 is ambiguous since there are two different notions of connectivity relevant to directed graphs. A directed graph is p a t h connected if given any two vertices a and 0, there is a sequence of vertices a = a , a i , . . . , a = 0 such t h a t for 0 < i < k, there is a n arrow a -> a . The graph is t o p o l o g i c a l ^ connected i f such a sequence of vertices exists and for 0 < i < k, there is either an arrow o n -> a or there is an arrow - o n - ( O f course, a path-connected graph is t o p o l o g i c a l l y connected, but i n general, the converse is false.) Fortunately, i n the presence of sufficient symmetry, the two concepts coincide.
Q k { m m

8.36. L e m m a . L e t Q be a finite d i r e c t e d g r a p h , a n d suppose that Q is topol o g i c a l l y connected. If t h e a u t o m o r p h i s m g r o u p of Q i s t r a n s i t i v e o n v e r t i c e s , then G ispath connected. P r o o f . It suffices to show t h a t if 0 -> a i n Q, then there is a p a t h (following arrows) from a to 0. L e t g be an a u t o m o r p h i s m of Q such t h a t 0-g = a , and suppose t h a t g has order n . Since g is a graph a u t o m o r p h i s m , we have a = 0-g -> a-g, and thus a-g - > a-g for a l l integers i. T h i s yields the path
i i+1

a > a - g > a-g and

> a-g ~

= a-g~

= 0,

this completes the proof.

In the situation of T h e o r e m 8.35, the group G acts on each o r b i t a l graph via graph automorphisms, a n d it is transitive on the vertex set ft. B y L e m m a 8.36, therefore, there is no real a m b i g u i t y i n the statement of 8.35. P r o o f of T h e o r e m 8.35. F i r s t , assume t h a t each nondiagonal o r b i t a l graph C ft is a n o n t r i v i a l block. Choose distinct

is connected, and suppose X

8D

261

points a and 0 G X , a n d let A be the o r b i t a l t h a t contains the pair (a,/?), so t h a t A is not the diagonal. B y assumption, the o r b i t a l graph corresponding to A is p a t h connected, a n d since X < ft, there must be a A - a r r o w leaving X . L e t 7 -> 8 be a A - a r r o w w i t h 7 e I and 6 X . T h e n b o t h ( a , /?) a n d (7, 5) lie i n A , and hence there exists g G G c a r r y i n g (a, (3) to (7, 5). N o w 7 G A and also 7 = a-g G A - # , so X a n d X - g are not disjoint. B u t X is a block, and thus 8 = 0-g G X - g = X . T h i s is a c o n t r a d i c t i o n , and hence G is p r i m i t i v e . Conversely, suppose that G is p r i m i t i v e , a n d fix a non-diagonal o r b i t a l A . T h e o r b i t a l graph of A can be decomposed into disjoint t o p o l o g i c a l ^ connected components. These components are p e r m u t e d b y the g r a p h aut o m o r p h i s m s induced by elements of G, a n d it follows t h a t each component is a block. B u t A is not the diagonal, and hence each point is j o i n e d by a A - a r r o w to some other point. A connected component X of the corresponding o r b i t a l graph thus contains more t h a n one point, and by p r i m i t i v i t y , it follows t h a t X = ft. T h e graph is thus t o p o l o g i c a l ^ connected, a n d hence it is p a t h connected too. F i n a l l y , we are ready use this graph-theoretic point of view to establish some results about suborbits and subdegrees of p r i m i t i v e actions. 8 . 3 7 . T h e o r e m . L e t G be a g r o u p a c t i n g p r i m i t i v e l y o n ft, a n d t h a t t h e subdegrees a r e 1 = m < m < < m, w h e r e r i s t h e r a n k . f o r l < i < r , w e h a v e m /m < m.
x 2 r l+1 l 2

suppose Then

P r o o f . L e t a G ft, and let A be an o r b i t a l w i t h | A ( a ) | = m . Since we can assume t h a t A is not the diagonal, the corresponding o r b i t a l g r a p h is p a t h connected, and so for each point 0 G ft, there is a p a t h (following A - a r r o w s ) from a to 0. W r i t e d{0) to denote the length of the shortest such path, so t h a t for example, d ( a ) = 0, and d{0) = 1 precisely w h e n 0 G A (a).
2

Now let i < r . l i m = m , there is n o t h i n g to prove, so we can assume that m > m i , and hence there is at least; one G - o r b i t of size exceeding m i . L e t 0 lie i n one of these large G - o r b i t s , a n d choose 0 so that d = d{0) is as small as possible. Observe that d > 1 because otherwise, 0 is either equal to a or else it lies i n A (a), and i n either case it lies i n a G - o r b i t of size at most m , w h i c h does not exceed m j . It follows t h a t there exists a point 7, where (7) = d - 1 and there is a A - a r r o w 7 /?.
x i + 1 l + 1 a Q Q 2

Now let X and Y be the unique orbitals such that ( a , 7) G X and (a,0) G Y. B y the choice of 0, we have \ Y { a ) \ > m , and so \ Y ( a ) \ > m . A l s o , we have \ X ( a ) \ < m i by the m i n i m a l i t y of d.
l l + 1

262

8. P e r m u t a t i o n

Groups

One can travel from a to 0 by first traversing the X - a r r o w from a to 7, a n d then the A - a r r o w from 7 to 0. Since G fixes a and induces automorphisms of b o t h the X - g r a p h and the A - g r a p h , it follows that one can travel from a to an a r b i t r a r y point i n the G - o r b i t of (3 by first traversing some X - a r r o w and then some A - a r r o w . In other words, every point i n Y { a ) can be reached i n this way.
a a

T h e number of X - a r r o w s leaving a is \ X ( a ) \ < m and the number of A - a r r o w s leaving each point i n ft is m . It follows that the number of different points that can be reached from a by first traversing an X - a r r o w a n d then a A - a r r o w is at most m m . W e now have
u 2 2 i

m + i and the result follows.

2

< \Y{<*)\

<

mm
2

i)

N o t e t h a t i f m = 1 i n T h e o r e m 8.37, then a l l G - o r b i t s have size 1, w h i c h means that G acts t r i v i a l l y on ft. T h u s if G is a p r i m i t i v e perm u t a t i o n group and a point stabilizer fixes a second point, then the point stabilizer is the t r i v i a l subgroup. B y p r i m i t i v i t y , however, the point stabilizer is a m a x i m a l subgroup, and thus \ G \ is prime. B u t of course, one does not really need a n y t h i n g as sophisticated as T h e o r e m 8.37 to o b t a i n this conclusion, w h i c h is P r o b l e m 8 B . 5 .
a

N e x t , we give another necessary c o n d i t i o n for a list of positive integers to be the subdegrees of a p r i m i t i v e action. 8 . 3 8 . T h e o r e m (Weiss). L e t G be a g r o u p a c t i n g p r i m i t i v e l y o n ft, a n d l e t n be t h e l a r g e s t s u b d e g r e e . If m i s a subdegree coprime ton , then m = l . It is convenient to introduce a little more notation. Suppose that a , fie ft, where, as usual, G is transitive on ft. T h e n the pair (a,0) lies i n some o r b i t a l A , and i n this situation, we have said that a -> (3 is a A - a r r o w . W h e n our p r i m a r y concern is the size of the G - o r b i t containing 0, we w i l l say t h a t a - 0 is an m-arrow, where m is this orbit size. N o t e t h a t
a

m = \ G

: G

n G \ = \Gp : G
0
a

n G \,
0

where the second equality holds because \ G \ = \ G p \ . If a -* 0 is a n m arrow, therefore, then 0 -* a is also an m-arrow. T h i s s y m m e t r y is also a consequence of the fact t h a t m = | A | / | f t | = | A | / | ^ | , where A ' is the o r b i t a l paired w i t h A .
/

T h e first conclusion of the following l e m m a is needed for our proof of T h e o r e m 8.38; the second and t h i r d w i l l not be used u n t i l later. 8 . 3 9 . L e m m a . L e t G be a g r o u p a c t i n g t r a n s i t i v e l y o n ft, a n d l e t a,0,j G ft. Suppose t h a t a -> 0 i s a n m - a r r o w a n d a -> 7 i s a n n - a r r o w , w h e r e m

8D

263

a n d n a r e c o p r i m e , a n d l e t u be t h e i n t e g e r a s s o c i a t e d w i t h t h e a r r o w s (3 - + 7 and-f^p. T h e following then hold. (a) u > n . (b) u d i v i d e s m n . (c) G G

7 Q

G G .
y 0

N o t e t h a t the subgroup products i n L e m m a 8.39(c) are subsets of G , but they need not be subgroups. Before presenting the proof, we briefly review a few elementary facts. L e t X and Y be subgroups of a finite group G. T h e n \ G \ > \ X Y \ = | X | | Y | / | X n Y | , and thus \ G : Y \ > \ X : X n Y \ . A l s o , \ G : X \ a n d \ G : Y\ divide \ G : X n Y\, a n d so if these indices are coprime, t h e n \ G : X \ \ G : Y\ divides \ G : A D Y | . T h e n | G : A | | G : y | < | G : X D Y | , and we deduce t h a t |G| < | X | | Y | / | X n Y\. E q u a l i t y thus holds i n a l l of these inequalities, and i n particular, |G : Y\ = \ X : X n Y\ and | G : X \ \ G : Y\ |G : X n y | . A l s o , G = A Y i n this case.

P r o o f o f L e m m a 8 . 3 9 . W r i t e A = G , B = Gp and C = G . T h e n \ A : A n B \ = m and \ A : A n G | = n , as indicated i n the diagram. Since m and n are coprime, it follows t h a t
a 7

n = \ A : A n C\ = \ A n B : ( A n C ) n ( A n B ) \ = \ A n B :( A n B ) nc\

< \ B : B D C \

264

8. P e r m u t a t i o n

Groups

p r o v i n g (a). A l s o , since \ A \ = \ B \ , we see that u = \ B : B n C\ divides \B :A n B n C \ = \A-. = \A: A n B n C \ ( A n B ) n ( A n C ) \

= \ A: A n B \ \ A : A n C \ = ran,

a n d this establishes (b). F i n a l l y , A = ( A n C ) ( A n B ) , and thus C A = C { Ar \ C ) { A r \ B ) a n d this is (c). = C { A r \ B ) C C B ,

P r o o f of T h e o r e m 8.38. L e t a G ft. A s s u m e that m ^ 1, and choose an o r b i t a l A so t h a t | A ( a ) | = m. T h e n A is not the diagonal, and hence by p r i m i t i v i t y and T h e o r e m 8.35, the corresponding o r b i t a l graph is connected. E v e r y A - a r r o w is an m-arrow, however, and hence there is a p a t h consisting of m-arrows from a to every other point i n ft. F o r 5 G ft, let d ( 8 ) denote the length of the shortest such path, a n d note that d ( a ) = 0. N o w choose (3 G ft so t h a t a - f3 is an n -arrow. O f course, an m-arrow joins (3 to each point i n A(/3), and since m ^ n , we w i l l have our desired c o n t r a d i c t i o n w h e n we prove that every arrow leaving (3 is a n n -arrow, or equivalents, that every arrow a r r i v i n g at 3 is a n n -arrow. N o w let 7 G ft. We show t h a t the arrow 7 -> (3 is an n -arrow by i n d u c t i o n on d = d ( j ) , and t h a t w i l l complete the proof. If d = 0, then 7 = a , and since a -+ 3 is an n -arrow, there is n o t h i n g further to prove i n this case. N o w suppose that d > 0, so that there exists a point 5 such that 5 -> 7 is an m-arrow and d ( 5 ) = d - 1. B y the inductive hypothesis, 8 -> /? is an n -arrow, and since m and n are coprime, it follows by L e m m a 8.37 t h a t 7 -> /? is a it -arrow, where u > n. B u t b y hypothesis, n is the largest subdegree, and thus u = n a n d the proof is complete. T h e following theorem about m u l t i p l e t r a n s i t i v i t y can also be proved by graph-theoretic techniques. 8.40. T h e o r e m ( M a n n i n g ) . Suppose ft, a n d l e t n be t h e l a r g e s t subdegree. s t a b i l i z e r acts 2 - t r a n s i t i v e l y o n some a n d G i s 3 - t r a n s i t i v e o n ft. G is a group acting primitively o n Assume that n > 3 a n d that a point s u b o r b i t of size n . Then n = |ft| - 1,

P r o o f . L e t a G ft. B y L e m m a 8.2, we must show t h a t G is 2-transitive on ft - { a } . B u t G acts 2-transitively on an orbit of size n i n ft, so it suffices to show t h a t n = |ft| - 1.
a a

8D

265

B y hypothesis, there exists an o r b i t a l A such t h a t G is 2-transitive on A (a) and \ A ( a ) \ = n . L e t X be an o r b i t a l (allowing X = A ) such that a n X arrow joins two distinct points of A (a). Since G induces automorphisms of the o r b i t a l graph corresponding to X a n d G is 2-transitive on A (a), it follows that every arrow j o i n i n g two distinct points of A (a) (in either direction) is a n X - a r r o w . In p a r t i c u l a r the o r b i t a l X is self-paired.
a a a

Let 0 G A ( a ) . T h e n X(f3) includes a l l of A ( a ) - {/?}, n - l . N o w n - 1 is relatively prime to n, a n d since n - 1 largest subdegree, it follows by T h e o r e m 8.38 t h a t n - 1 is T h u s \X(0)\ = n , and so there is a unique point 7 G X { 0 )
a a

and so \X(0)\ > > 1 and n is the not a subdegree. with 7 0 A ( a ) .

Since G is 2-transitive on A (a), we know t h a t [ G ) p acts transitively on the n - 1 points of A ( a ) - { 0 } = X(/3) - {7}. A l s o , since (G )p is contained i n Gp, it stabilizes the set X { 0 ) , a n d it follows that (G )p fixes 7. T h e n { G ) f 3 C ( G ) , and since ( G ^ is transitive o n X(/3) - {7}, we see t h a t Gp is 2-transitive on X { 0 ) .
a a a / 3 7

Now n > 3, so there are at least two points i n the set X { 0 ) - {7} = A ( a ) - { p } , and since these points lie i n A ( a ) , the arrow j o i n i n g t h e m must be an A - a r r o w . B u t Gp is 2-transitive on X(0), a n d so by the reasoning we used before, every two distinct points i n X { p ) are j o i n e d by an A - a r r o w . Furthermore, because G is transitive on ft and defines automorphisms of the o r b i t a l g r a p h corresponding to X , it follows t h a t for a n a r b i t r a r y point 8 G ft, every two distinct points of X(8) are j o i n e d b y a n X - a r r o w . We now complete the proof by showing t h a t X { p ) u { P } = ft, so that n = |ft| - 1 , as wanted. If this is false, then since the o r b i t a l graph corresponding to X is connected, there exists an A - a r r o w 8 - e, where 8 G X { P ) U { P } and e 0 X { P ) U { p } . T h e n e G X { 8 ) , and since e 0 A(/3), we have 5^0. Then 5 G X { 0 ) , and hence /? G X ( 5 ) because A is self paired. A l s o , e 7^ 0, and since /? and e lie i n X(<5), it follows by the previous paragraph that 0 -> e is an A - a r r o w . T h i s is a contradiction, however, since e#X(0). E v e n for i m p r i m i t i v e groups, these graph-theoretic techniques can be used to establish some necessary conditions o n the set of subdegrees. T o state our next result (which is a theorem of C . Praeger and the author) we introduce yet another graph. G i v e n a set D of positive integers, the (undirected) c o m m o n - d i v i s o r g r a p h Q ( D ) is constructed as follows. T h e vertex set of Q { D ) is D , and distinct integers o, b G D are joined by an edge precisely when they are not relatively p r i m e . N o t e that i f 1 G D , then no edge is incident w i t h 1, and so {1} is a connected component of Q ( D ) . 8 . 4 1 . T h e o r e m . L e t D be t h e set of d i s t i n c t subdegrees f o r some transit i v e g r o u p a c t i o n . T h e n t h e c o m m o n - d i v i s o r g r a p h G { D ) has a t m o s t t h r e e connected components, including { ! } .

266

8. P e r m u t a t i o n

Groups

T h u s , for example, the set { 1 , 2 , 3 , 4 , 5 } cannot be the full set of subdegrees for a transitive group action because the corresponding commondivisor g r a p h has four connected components: {1}, { 2 , 4 } , {3} and {5}. Unfortunately, we need to introduce still more n o t a t i o n . L e t G be transitive on ft, a n d let m be a subdegree. R e c a l l that i f a , 0 G ft, we declared a -> (3 to be an m-arrow if 0 lies i n a G - o r b i t of size m , or equivalently, if a lies i n a G o r b i t of size m . F o r a G ft, w r i t e [ a ] to denote the subset of ft consisting of a a n d all points that can be reached from a along a p a t h of m-arrows. I n other words, [ a ] is the connected component containing a i n what might be called the m - g r a p h on ft, denned by the m-arrows. Since each element of G carries m-arrows to m-arrows, we see that G acts as aut o m o r p h i s m s of the m - g r a p h , and thus it permutes the (necessarily disjoint) connected components. It follows that [ a ] is a block.
a r m m m

N o w write K (a)
m

= ( G p | 0 G [a] >,
m

and K ( k and
m m

observe that G C K { a ) . It should be clear t h a t if g G G, we have a - g ) = K { a ) 3 , and thus since G acts t r a n s i t i v e l y on ft, the index = \ K ( a ) : G \ is independent of a . O f course, k does depend on m , we stress that it is denned only when TO is a subdegree.
a m m m a m a a 0

If a - + 0 is a 1-arrow, then G fixes 0, and thus G = G . It follows that K x i a ) = G , and hence k = 1. If a /? is a n m-arrow and TO > 1, on the other hand, then G does not fix 0, and thus G ^ G and # ( a ) 5 ( G , G/j) > G . If m > 1, therefore, we have k > 1.
a Y a a 0 m a Q m

8 . 4 2 . T h e o r e m . L e i G be a g r o u p a c t i n g t r a n s i t i v e l y o n ft. Suppose that m < n a r e subdegrees, a n d assume t h a t every subdegree i s c o p r i m e t o a t least one of m o r n . T h e n (a) k (b) k
m

divides k

and

divides n .

P r o o f . F i x 7 G ft and consider the nonempty set S of a l l points a G ft such that a > 7 is a n n-arrow. If a 6 5 and a > 0 is an m-arrow, we argue t h a t 0 e S. W e k n o w that 0 - 7 is a u-arrow for some subdegree n , and our goal is to show t h a t u = n . B y L e m m a 8.39(a), we have u > n , and if u > n , t h e n of course also u > m , a n d so u cannot divide either n or m . B u t L e m m a 8.39(b) tells us t h a t u divides m n , and so if u is coprime to either m or n , it w o u l d have to divide the other, w h i c h is not the case. T h i s u is not coprime to either m or n , w h i c h is contrary to the hypothesis. It follows t h a t u = n , and thus 0 G 5 , as claimed. N o w fix a S. Since the set [ a ] is connected v i a m-arrows, it follows by the argument of the previous paragraph that [ a ] C S. T h e n [ a ] C [ ] ,
m m m 7 n

8D

267

a n d hence K

{ a ) C K {j).
n

Since \ G \ = | G | ,
a 7 w 7 7

we have

fc

|jr ( ):G | |ATm(a) : G a |

Km

w h i c h is an integer. T h i s establishes (a). If a /? is an m-arrow, we have G G C G G by L e m m a 8.39(c). Since /? a is also an m-arrow a n d we have shown t h a t 0 G S, we can interchange the roles of a and 0 to deduce the reverse containment, and thus G G = G G . Since [ a ] is connected v i a m-arrows, this reasoning shows that for a l l 5 G [ a ] , the sets G G are equal, and thus G G is invariant under right m u l t i p l i c a t i o n by each of the subgroups G . It follows that G y G is invariant under right m u l t i p l i c a t i o n by K { a ) , w h i c h is the subgroup is generated by these G . W r i t i n g K = K ( a ) a n d noting t h a t G C K , we have
1 a 7 / 3 7 Q 7 0 m m 7 5 7 a s a m s m a

G G
7

= G ^ G K = GyK ,
a

a n d thus \ -y\\ \ |G Then \ G n

= a 7 7 Q G G

_ ,^
| G 7

n G q |

_ , l

n G

, _ ^
K i

|G ||K|
7

n G | a

- ] G ^ K \ \ G \
= | C ? 7 : G

: G n G | = | K : G n K \ , and this yields

7

|GQ,:G nG |
a 7

\K-.GjC\K\
=

\ K : G

\ K:G |
a 7

|G n K|
7 Q

>'

where the first equality holds because lies i n a G - o r b i t of size n , and the last one holds because \ G \ = | G | . Since the q u a n t i t y on the right is an integer, it follows that k divides n , and this completes the proof.
a 7 m

T h e o r e m 8.41 follows easily from T h e o r e m 8.42(b). P r o o f o f T h e o r e m 8 . 4 1 . Suppose t h a t u,v,w G D lie i n three different connected components of the c o m m o n - d i v i s o r g r a p h Q { D ) . In p a r t i c u l a r u, v and w are distinct, and so we can assume t h a t u < v a n d u < w. Since u a n d v lie i n different components, every subdegree is coprime to at least one of u or v, a n d so k divides v by T h e o r e m 8.42(b). Similarly, k divides w, a n d since v and w are coprime, it follows t h a t k = 1, a n d thus u = 1. T h i s shows that there do not exist three connected components different from {1} i n Q { D ) , a n d so there are at most three components i n t o t a l .
u u u

U s i n g T h e o r e m 8.42(a) we can get a d d i t i o n a l i n f o r m a t i o n about connected components of Q { D ) . 8 . 4 3 . T h e o r e m . L e t D be t h e set of subdegrees assume that the c o m m o n - d i v i s o r G { D ) actually nents: {1}, A a n d B . If B c o n t a i n s t h e l a r g e s t a l l a e A a n d be B . A l s o , every two integers i

the

for a transitive action, and has t h r e e c o n n e c t e d c o m p o subdegree n , then a < b for n B a r e j o i n e d by a n edge.

268

8. P e r m u t a t i o n G r o u p s

P r o o f . L e t a G A and b G B , and suppose that a > b. T h e n fc divides a by 8.42(b), a n d also k divides fc by 8.42(a). Furthermore, since a A and n G B , we have a < n, a n d thus k divides n by 8.42(b). Since k divides k , we see t h a t A; divides n, and since it also divides a a n d fc > 1, it follows that a and n are not coprime. T h i s is a contradiction, however, since a and n are i n different connected components of the common-divisor graph.
6 b a a b a 6 b

If a G A and u, v G J3, then a < u a n d a < b, a n d thus fc divides b o t h u and v. Since k > 1, we conclude t h a t u and v are joined by an edge.
a a

We saw previously that the set { 1 , 2 , 3 , 4 , 5 } cannot be the full set of subdegrees of a transitive action because it does not satisfy the conclusion of T h e o r e m 8.41. T h e connected components for the set {1, 2, 3,4}, however, are {1}, {2,4} and {3}, and so this set is not eliminated by T h e o r e m 8.41. B y T h e o r e m 8.43, however, {1, 2, 3,4} cannot be a subdegree set because i n the n o t a t i o n of t h a t theorem, n = 4, a n d thus A = {3} a n d B = { 2 , 4 } , and we do not have a < b for a l l a G A and b G B . We close w i t h an a p p l i c a t i o n of this m a t e r i a l outside of the realm of transitive p e r m u t a t i o n groups. 8.44. C o r o l l a r y . L e t A a c t o n G v i a a u t o m o r p h i s m s , w h e r e A a n d G a r e finite g r o u p s , a n d l e t D be t h e set of sizes of t h e A - o r b i t s o n G. Then t h e c o m m o n - d i v i s o r g r a p h Q { D ) has a t most three connected components i n c l u d i n g {1}. A l s o , if t h e r e a r e t h r e e c o m p o n e n t s {1}, A a n d B , w h e r e B c o n t a i n s t h e l a r g e s t member of D , t h e n a < b f o r a l l a G A a n d be B , a n d every t w o i n t e g e r s i n B a r e j o i n e d by a n edge.

P r o o f . L e t T = G x A , a n d view A and G as subgroups of T. T h e n T acts transitively on the set ft of right cosets of A i n T, a n d A is the stabilizer of the "point" A e ft. T h e action of A on ft - { A } is p e r m u t a t i o n isomorphic to the action of A on the nonidentity elements of G, a n d thus the action of A o n ft is p e r m u t a t i o n isomorphic to its original action o n G. In particular, D is the set of subdegrees for the action of Y on ft, a n d hence Theorems 8.41 and 8.43 apply.

A n i m p o r t a n t special case of C o r o l l a r y 8.44 is where A = G, acting by conjugation. In t h a t situation, D is just the set of class sizes of G, a n d it follows t h a t the set of class sizes of a n a r b i t r a r y finite group satisfies the conclusion of C o r o l l a r y 8.44, and i n particular, the common-divisor graph on this set has at most three connected components. N o t surprisingly, more can be said i n this case t h a n seems to be available from the general theory we have presented.

Problems 8D

269

Problems

8D
2 r

8D.1. L e t 1 = m i < m < < m be the subdegrees for a p r i m i t i v e action of rank r . If m = 2, show that m = 2 for a l l i > 2.
2

8 D . 2 . Show t h a t a p r i m i t i v e p e r m u t a t i o n group of degree 8 must be d o u b l y transitive. 8 D . 3 . L e t G be a transitive p e r m u t a t i o n group of r a n k r , a n d suppose that the largest subdegree is n . P r o v e t h a t \ G \ is b o u n d e d by some function of r and n . 8D.4. L e t m be a subdegree for a transitive a c t i o n of G on ft, a n d let be the integer denned just before T h e o r e m 8.42. (a) Show that k
m m

> m.

(b) If k = m a n d a -> 0 is a n m - a r r o w for a , 0 G ft, show that is a subgroup. (c) If m > 1 a n d G is p r i m i t i v e on ft, show t h a t k
m

> m.

8 D . 5 . L e t G act t r a n s i t i v e l y on ft, a n d suppose the rank is 3. If the subdegrees are 1 < m < n , where m a n d n are coprime, show that m + 1 must divide n . H i n t . Observe that k
m

must divide 1 + m + n .

8 D . 6 . L e t G be a p r i m i t i v e p e r m u t a t i o n group, a n d suppose that some p r i m e number p is a subdegree. P r o v e that p does not divide the order of a point stabilizer, a n d deduce t h a t p divides no subdegree.
2 2

Hint. I f X c y

a n d \Y : X \ = p , p r o v e t h a t Q P ' { X ) < Y.

Chapter 9

More on Subnormality

9A In this chapter, we present a few more applications of s u b n o r m a l i t y theory, and also a pretty result w i t h i n the theory for w h i c h no a p p l i c a t i o n seems to be k n o w n . W e begin w i t h a discussion of H . B e n d e r ' s "generalized F i t t i n g subgroup". T h i s characteristic subgroup a n d its associated "components" have become s t a n d a r d tools i n group theory, a n d they have played an essent i a l role i n the classification of simple groups. R e c a l l that the F i t t i n g subgroup of a group G is the unique largest n o r m a l nilpotent subgroup. in G . If G is solvable, it follows by P r o b l e m 3 B . 1 4 that the F i t t i n g subgroup F ( G ) is large enough to c o n t a i n its own centralizer If we make the weaker a s s u m p t i o n t h a t G is p-solvable for some p r i m e p , a n d we assume i n a d d i t i o n t h a t C y ( G ) = 1, it is easy to see t h a t F ( G ) = O ( G ) , a n d it follows by the H a l l - H i g m a n L e m m a 1.2.3 that i n this
p

case too, F ( G ) contains its centralizer i n G . B u t i n general, of course, F ( G ) can fail to contain its centralizer. T h i s c e r t a i n l y happens, for example, if G is a n o n a b e l i a n simple group because i n t h a t case, F ( G ) = 1. In this section, we define the generalized F i t t i n g subgroup F * ( G ) , a n d we prove that this characteristic subgroup always contains its centralizer i n G. I n general, F * ( G ) 3 F ( G ) , a n d we shall see that i n cases where the F i t t i n g subgroup contains its centralizer, F * ( G ) = F ( G ) . I n other words, if no augmentation of the F i t t i n g subgroup is necessary i n order for it to contain its centralizer, t h e n no a u g m e n t a t i o n occurs w h e n we construct the generalized F i t t i n g subgroup. W e begin work by recalling that a group G is said to be perfect if G' = G.

271

272

9. M o r e o n S u b n o r m a l i t y

9.1. L e m m a . L e t G be a g r o u p , a n d suppose that G / Z { G ) is simple. Then G / Z { G ) i s n o n a b e l i a n , a n d G' i s p e r f e c t . A l s o G ' / Z { G ' ) i s i s o m o r p h i c t o the simple group G / Z { G ) .

P r o o f . W r i t e Z = Z ( G ) . If the simple group G / Z is abelian, it must be cyclic, and thus G is abelian and Z = G, w h i c h is a c o n t r a d i c t i o n . T h u s G / Z is a nonabelian simple group, as claimed. In particular, G is not solvable, and thus G'" ^ 1, so G" is not abelian, and hence G" Z. Now Z is a m a x i m a l n o r m a l subgroup of G and G" Z, and thus G " Z > Z, a n d we conclude that G " Z = G. T h e n G / G " = G " Z / G " =i Z / { Z n G " ) , w h i c h is abelian. T h e n G / G " is abelian, a n d G' C G". T h e reverse containment is obvious, so G" = G', and G' is perfect. F i n a l l y , G ' / { Z n G ' ) = G ' Z / Z = G / Z is simple. It follows that Z n G ' is a m a x i m a l n o r m a l subgroup of G\ and since G' is nonabelian, we see that Z n G ' must be the full center of G'. T h u s G ' / Z { G ' ) = G / { Z n G') G/Z, and the proof is complete.

A group H is said to be q u a s i s i m p l e if H / Z ( H ) is simple and H is perfect. B y L e m m a 9.1, therefore, if G / Z ( G ) is simple, then G' is quasisimple. T o produce examples of quasisimple groups, we observe first t h a t a l l nonabelian simple groups are quasisimple. A l s o , the groups S L { n , q ) are quasisimple i f n > 3 or n = 2 and q > 3, and most of these groups are not simple. (See Theorems 8.32 a n d 8.33.) Of course, if G is quasisimple, there is a unique nonabelian simple group G / Z ( G ) associated w i t h it. A t first, it may seem that for each nonabelian finite simple group, there should be m a n y (perhaps even infinitely many) associated quasisimple groups. I n fact, there are o n l y finitely m a n y (and usually o n l y a few). A l t h o u g h we w i l l not present a proof here, it is a fact that for each nonabelian simple group S, there is a unique largest quasisimple group G (called the Schur representation group of S) such that G / Z ( G ) = S. A l l of the quasisimple groups associated w i t h S have the form G/Y, where G is the representation group of S and Y C Z ( G ) . (The abelian group Z(G) is the Schur m u l t i p l i e r of S, and we refer the reader to a slightly more detailed discussion of Schur multipliers i n C h a p t e r 5.) F o r example, if S is the alternating group A w i t h n > 5, then unless n = 6 or n = 7, the Schur m u l t i p l i e r has order 2, and so there are exactly two quasisimple groups associated w i t h A i n this case: the simple group A itself, and the corresponding representation group, w h i c h has order 2 \ A \ . If n is 6 or 7, the Schur m u l t i p l i e r of A is cyclic of order 6, and so there are exactly four associated quasisimple groups i n these two cases, w i t h orders \ A \ , 2 \ A \ , Z \ A \ and 6 | A | .
n n n n n n n n n

9A

273

9.2. L e m m a . L e t G be q u a s i s i m p l e . If N i s a p r o p e r n o r m a l s u b g r o u p of G, t h e n N C Z ( G ) . A l s o , every n o n i d e n t i t y h o m o m o r p h i c i m a g e of G i s quasisimple. P r o o f . W r i t e Z = Z ( G ) , so that Z is a m a x i m a l n o r m a l subgroup of G , and let TV < G w i t h N < G. If i V % Z , t h e n N Z > Z , a n d thus 7VZ = G by the m a x i m a l i t y of Z . T h e n G / i V = T V Z / i V = Z / ( i V n Z ) is abelian, a n d so G = G' C N < G . T h i s c o n t r a d i c t i o n shows t h a t N C Z . To prove the second_assertion, we_show t h a t G = G / / V is quasisimple. W e have ( G ) ' = G ^ = _ G , and thus G is perfect, as required. A l s o , since JV C Z , we have G / Z ^ G / Z , w h i c h is simple and nonabelian. Since Z C Z ( G ) and the factor group G / Z is a n o n a b e l i a n simple group, it follows t h a t Z must be the full center of G , and this completes the proof. A s u b n o r m a l quasisimple subgroup of a n a r b i t r a r y finite group G is called a c o m p o n e n t of G . (Of course, it can h a p p e n t h a t G has no components; this is the s i t u a t i o n if G is solvable, for example.) Observe t h a t if H is a component of G , then it is also a component of every subgroup of G i n w h i c h it happens to be contained. 9.3. L e m m a . L e t N be a m i n i m a l n o r m a l s u b g r o u p of a finite g r o u p G, a n d suppose t h a t H i s a c o m p o n e n t of G w i t h H % N . T h e n [ N , H ] = 1. P r o o f . F i r s t , H D N < H , a n d of course H n N < H . B y L e m m a 9.2, therefore, H D N C Z { H ) . N o w H G a n d N is m i n i m a l n o r m a l i n G , and so by T h e o r e m 2.6, we k n o w t h a t N C N { H ) , a n d thus [AT, i f ] C H . A l s o , [TV, #] C N since N < G, and hence [TV, #] C # n N C Z ( / f ) . T h e n [AT, H , H } = 1 a n d [#, A/", P ] = [A", H , H } = 1 , and so P , /V] = 1 by the three subgroups l e m m a . Since H = H ' , we have [ i i , A/] = [ H , H , N ] = 1, and the proof is complete.
G

9.4. T h e o r e m . L e t H a n d K be d i s t i n c t c o m p o n e n t s T h e n [ H , K ] - 1.

of a

finite

group

G.

P r o o f . W e proceed by i n d u c t i o n o n | G | . If b o t h H and K are contained i n some proper subgroup A of G , then H and K are distinct components of X . B y the i n d u c t i v e hypothesis a p p l i e d i n X , it follows t h a t [ H , AT] = 1, as required, and so we can assume that no proper subgroup of G contains b o t h H and K . If G is simple, then since H a n d K are n o n t r i v i a l a n d s u b n o r m a l , we have H = G = K , w h i c h is not the case. T h u s G is not simple, and since it is certainly n o n t r i v i a l , G has a m i n i m a l n o r m a l subgroup N . A l s o , N < G, and thus N does not contain b o t h H and AT. If one of these components, say K , is contained i n N , then H % N , and thus [ H , K ] C [ H , N ] = 1, by

274

9. M o r e o n S u b n o r m a l i t y

L e m m a 9.3, and we are done i n this case. W e can assume, therefore, that for every m i n i m a l n o r m a l subgroup N , we have H N and K % N . Let G G/N, where N is m i n i m a l n o r m a H n G, and observe that H_ and K_ are nonidentity s u b n o r m a l subgroups of G. B y Lemma_9.2, b o t h H and K are quasisimple, and so they are components of G. 1iH_ K , then by the i n d u c t i v e hypothesis applied i n the group G , we have [ H , K ) = 1, and thus [ H , K ] C N . B u t [TV, H ) = 1 by L e m m a 9.3, and thus [ H , K , H ] = 1 and [ K , H , H ] = [ H , K , H ] = 1. B y the three subgroups l e m m a , therefore, we have 1 = [ H , H , K ] = [#,AT], where the second equality holds because H = H'. W h a t remains is the case H = K , or equivalently, H N = K N , and we can assume that this equation holds for every choice of a m i n i m a l n o r m a l subgroup N of G. Since H N contains b o t h H and K , it follows that H N = G. B y T h e o r e m 2.6, the m i n i m a l n o r m a l subgroup N normalizes H , and thus H < G , and s i m i l a r l y K < G , and hence [ H , K \ C # n K . If AT] > 1, we c a n choose our m i n i m a l n o r m a l subgroup N so t h a t N C H n K , and thus Tf = ZT/V - K A ^ = a n d we have a c o n t r a d i c t i o n . It follows from T h e o r e m 9.4 that the components of a finite group G normalize each other, a n d so each component is n o r m a l i n the subgroup generated by a l l of t h e m . T h i s subgroup, denoted E ( G ) , is called the l a y e r of G . T h e layer of G is thus the p r o d u c t of the components of G , and of course, it is characteristic i n G . (Of course, if G has no components, then E ( G ) - 1.) In order to discuss the properties of the layer, it is convenient to introduce a another type of group that generalizes nonabelian simple groups. W e say t h a t a group G is s e m i s i m p l e if it is a product of nonabelian simple n o r m a l subgroups. (Unfortunately, not a l l authors agree on this definition, so readers should exercise caution.) A s the next l e m m a shows, semisimple groups are a c t u a l l y d i r e c t products of nonabelian simple groups. 9.5. L e m m a . Suppose t h a t a finite g r o u p G i s t h e p r o d u c t of t h e members of some c o l l e c t i o n X of n o n a b e l i a n s i m p l e n o r m a l s u b g r o u p s . T h e n t h e p r o d u c t i s d i r e c t , a n d X i s t h e set of a l l m i n i m a l n o r m a l s u b g r o u p s o f G . P r o o f . T h e members of X are clearly m i n i m a l n o r m a l subgroups of G , and thus if S E X and N is a m i n i m a l n o r m a l subgroup of G different from S, then N n S = 1, and hence S centralizes N . It follows t h a t the product of all members of X different from N centralizes N . We c a n certainly assume that X is nonempty. L e t T G X , and let AT be the p r o d u c t of a l l other members of X , so that TAT = G , and by the

9A

275

result of the first paragraph, K centralizes T. Since T is nonabelian and simple, it has t r i v i a l center, and thus T n K C Z ( T ) = 1, and it follows that G = T K = T x K . N o w K is the p r o d u c t of the members of the collection X - { T } , so w o r k i n g by i n d u c t i o n on \X\, it follows that this product is direct, and thus G = T x K is the direct p r o d u c t of the members of X . F i n a l l y , if N is m i n i m a l n o r m a l i n G a n d N X , then by the result of the first paragraph, N is centralized by T\ * = G, and thus N C Z ( G ) . B u t G is a direct product of groups h a v i n g t r i v i a l centers, so Z ( G ) = 1, and this contradiction shows that X contains a l l m i n i m a l n o r m a l subgroups of G . N e x t we show how semisimple groups arise n a t u r a l l y i n finite group theory. 9 . 6 . L e m m a . L e t N be a m i n i m a l n o r m a l s u b g r o u p Then either N is abelian o ri t is semisimple. of a finite group G.

P r o o f . L e t S be a m i n i m a l n o r m a l subgroup of N . If S is abelian, then S C F { N ) , and thus F(7V) is a n o n t r i v i a l n o r m a l subgroup of G contained in N . Since N is m i n i m a l n o r m a l i n G , we have F ( i V ) = N , and thus N is nilpotent. B u t then Z { N ) is a n o n t r i v i a l n o r m a l subgroup of G contained in N , and thus Z { N ) = N and N is abelian. We can now assume that S is nonabelian, and we argue that S is simple. If K < S, then K G , and thus by T h e o r e m 2.6, the m i n i m a l n o r m a l subgroup N of G normalizes K , a n d thus K < N . B u t K C S and 5 is m i n i m a l n o r m a l i n N . T h u s either K = 1 or K = S, and this shows t h a t 5 is simple, as claimed. We can now assume t h a t S is a nonabelian simple group. T h e n each conjugate of S i n G is a nonabelian simple group t h a t is n o r m a l i n N , and so the product S of these conjugates is semisimple. A l s o , 1 < S < G and S C N , and it follows by the m i n i m a l i t y of N t h a t S = N . T h u s N is semisimple, as wanted.
G G G G

It follows by L e m m a 9.5 that a nonabelian m i n i m a l n o r m a l subgroup of a n a r b i t r a r y finite group must be a direct p r o d u c t of nonabelian simple groups. We r e t u r n now to the layer, E ( G ) . finite group, and write Z =

9.7. T h e o r e m . L e t E = E ( G ) , w h e r e G i s a Z { E ) . The following then hold. (a) E ' = E . (b) E / Z is semisimple.

(c) [ E , M] = 1 f o r every

solvable n o r m a l subgroup M

ofG.

276

9. M o r e o n S u b n o r m a l i t y

P r o o f . L e t be the set of components of G , so t h a t E = \[. If H , t h e n H = H ' C E ' , and thus E ' contains a l l of the members of , and we haveE'Z2T\ = E , proving (a). Of course, [ Z { H ) , H] = 1, and if K w i t h K ^ H , t h e n \ Z { H ) , K ] C [If, AT] = 1 by T h e o r e m 9.4. T h u s Z ( H ) centralizes a l l members of , and hence it is central i n E . T h e n Z { H ) C J? n Z , and since the reverse containment is clear, we have Z { H ) = H n Z for a l l components # . N o w write = E / Z . If H , t h e n I f H / ( H n Z ) = H / Z ( H ) is a nonabelian simple group. A l s o ,

is a product of nonabelian simple n o r m a l subgroups, and thus E is semisim ple, proving (b). F i n a l l y , let M < G b e solvable. T h e n M n E is a n o r m a l subgroup of the semisimple group E . B y L e m m a 9.5, the m i n i m a l n o r m a l subgroups of E are nonabelian simple groups, and i n particular they are nonsolvable. It follows that the solvable n o r m a l subgroup M ( l E contains no m i n i m a l n o r m a l subgroup of E , a n d thus M D E = 1, and M D E C Z . T h e n [ M , ] C M n E a = Z ( ) , and hence [ M , E , E] = 1. A l s o , [ E , M , E] = [ M , E , E] = 1, and hence by the three subgroups l e m m a , 1 = [ E , E , M ] = [ E , M [ , where the second equality follows by (a). T h i s completes the proof. F i n a l l y , we define the generalized F i t t i n g s u b g r o u p of a finite group G by the formula F * ( G ) = F ( G ) E ( G ) . 9 . 8 . T h e o r e m . F o r every finite g r o u p G, w e h a v e F * ( G ) D C ( F * ( G ) ) .
G

P r o o f . L e t C = C ( F * ( G ) ) , and write Z = C n F * ( G ) , to show that Z = C. If this is false, t h e n Z < C, a n d n o r m a l i n G , we can choose a m i n i m a l n o r m a l subgroup M C C. Since Z C F * ( G ) , we see t h a t C centralizes Z , it follows t h a t Z C Z ( M ) .
G

so that our goal is since Z and G are M / Z of G / Z w i t h a n d since M C C ,

N o w M / Z is either abelian or semisimple by L e m m a 9.6. If M / Z is abelian, t h e n M is nilpotent, and so M C F ( G ) C F * ( G ) , and thus M C G n F * ( G ) = Z , w h i c h is a contradiction. T h u s M / Z is semisimple, and we let S/Z be a m i n i m a l n o r m a l subgroup of M / Z . T h e n S/Z is a nonabelian simple group by L e m m a 9.5, and it follows that Z is the full center of S. L e m m a 9.1 now applies, and we deduce that 5 ' is quasisimple. A l s o , S' G , and so 5 ' is a component of G , and S' C E ( G ) C F * ( G ) . T h e n 5 ' C C n F * ( G ) = Z , and thus S/Z is abelian. T h i s is a contradiction, and the proof is complete.

Problems 9 A

277

9.9. C o r o l l a r y . L e t G be a

finite
G

g r o u p . T h e n F*(G) DF(G), a n d equality

h o l d s if a n d o n l y i / F ( G ) 2 C ( F ( G ) ) . P r o o f . T h e containment is obvious. A l s o , since F * ( G ) contains its centralizer i n G , it follows that if F * ( G ) = F ( G ) , t h e n F ( G ) contains its centralizer. Conversely, suppose that F ( G ) 5 C ( F ( G ) ) . Since F ( G ) is solvable, E ( G ) C C ( F ( G ) ) by T h e o r e m 9.7(c), a n d thus E ( G ) C F ( G ) , a n d hence F * ( G ) = F ( G ) E ( G ) = F ( G ) , as required.
G G

Problems

9A

9 A . 1 . Suppose t h a t F * ( G ) C H C G . Show t h a t E ( H ) = E ( G ) . 9A.2. Show t h a t the socle of an a r b i t r a r y finite group is the direct product of a n abelian group w i t h a semisimple group. 9A.3. L e t N < G, where G is semisimple. Show t h a t N is the product of

those m i n i m a l n o r m a l subgroups of G t h a t it contains. H i n t . W r i t e G = U x V, where U is the p r o d u c t of the m i n i m a l n o r m a l subgroups of G t h a t are contained i n N . Show t h a t AT n V = 1. 9 A . 4 . L e t E = E ( G ) and suppose t h a t N < E . Show t h a t N = MY, where M is the product of a l l components of G t h a t are contained i n N a n d Y = MnZ(). 9 A . 5 . Suppose t h a t H < G and that C is a component of G not contained in H . Show that [ H , C ] = 1. H i n t . Use P r o b l e m 9 A . 4 to show that \ ( H n E ) , C] = 1, where E = E ( H ) . 9 A . 6 . L e t H < G , and suppose that H D C ( H ) .
G

Show t h a t H D E ( G ) .

9A.7. L e t N be a nonabelian m i n i m a l n o r m a l subgroup of G , and observe t h a t L e m m a s 9.5 a n d 9.6 i m p l y that N is a direct p r o d u c t of a collection X of nonabelian simple groups. P r o v e t h a t G acts t r a n s i t i v e l y by conjugation the set X . 9 A . 8 . A group G is characteristically simple i f it is n o n t r i v i a l and its only characteristic subgroups are 1 and G . If G is characteristically simple, show that G is a direct product of isomorphic simple groups. H i n t . Consider G x A u t ( G ) .

278

9. M o r e o n S u b n o r m a l i t y

9B If G is a n a r b i t r a r y group, then G / Z ( G ) is n a t u r a l l y isomorphic to the group of inner automorphisms I n n ( G ) . If Z ( G ) = 1, therefore, G I n n ( G ) , and so we can identify G w i t h I n n ( G ) v i a the n a t u r a l isomorphism, and this embeds G as a subgroup of A u t ( G ) . In fact, I n n ( G ) < A u t ( G ) , and so a group w i t h t r i v i a l center is n a t u r a l l y embedded as a n o r m a l subgroup of its a u t o m o r p h i s m group. ( A l l of this is routine, but for completeness, the details are given i n L e m m a 9.11 below.) It is easy to show (and we w i l l do this too i n L e m m a 9.11) t h a t i f Z ( G ) = 1, then also Z ( A u t ( G ) ) = 1, and so we c a n repeat the process and embed A u t ( G ) as a n o r m a l subgroup of its a u t o m o r p h i s m group. C o n t i n u i n g like this, we o b t a i n the a u t o m o r p h i s m tower associated w i t h G : G < Aut(G) < Aut(Aut(G)) < Aut(Aut(Aut(G))) < . One of the earliest applications of H . W i e l a n d t ' s s u b n o r m a l i t y theory was his beautiful a u t o m o r p h i s m tower theorem, w h i c h asserts that the automorp h i s m tower associated w i t h a finite group h a v i n g t r i v i a l center contains only finitely m a n y different groups. 9.10. Z(G) Gi. In this section, we present a proof of the a u t o m o r p h i s m tower theorem based on ideas of E . S c h e n k m a n and M . Pettet (and, of course, of W i e l a n d t ) . Before we begin to develop the necessary theory, we m e n t i o n that it does not seem to be k n o w n whether or not the hypothesis that G has t r i v i a l center is really necessary i n T h e o r e m 9.10. W i t h o u t this assumption, we do not have n a t u r a l embeddings of G i n G , so there is no a u t o m o r p h i s m tower, but nevertheless, it might be true t h a t there are just finitely m a n y different groups G . F o r example, i f G is dihedral of order 8, it happens that A u t ( G ) ^ G , and so there is only one group G i n this case.
{ l + l { %

Theorem = 1.

( W i e l a n d t ) . L e t G be a finitely many

finite

group,
{

a n d assume
l

that Then the

W r i t e G i = G, a n d f o r i > 1, l e t G

= Aut(G _i). groups among

up t o i s o m o r p h i s m , there are only

different

In the s i t u a t i o n of T h e o r e m 9.10, where Z ( G ) = 1, we have remarked (but we have not yet proved) that the subgroups G a l l have t r i v i a l centers, and so there is a n a t u r a l embedding G < G . It follows t h a t G is s u b n o r m a l in each of the groups G and this, of course, shows w h y s u b n o r m a l i t y theory is relevant here.
t i + 1 u

If we accept W i e l a n d t ' s theorem for the moment, and we let G be the largest of the groups G then clearly G = G . Since G = Aut(G ) and G is identified w i t h I n n ( G ) , the equation G = G tells us t h a t every a u t o m o r p h i s m of G is inner. W e mention that a group G w i t h Z ( G ) = 1 and
r u r r + 1 r + 1 r r r r r + 1 r

9B

279

A u t ( G ) = I n n ( G ) is said to be c o m p l e t e , a n d so it follows from W i e l a n d t ' s theorem t h a t every finite group w i t h a t r i v i a l center can be s u b n o r m a l l y embedded i n a complete group. Conversely, if one of the groups G i n the a u t o m o r p h i s m tower of G is complete, t h e n G = I n n ( G ) = A u t ( G ) = G + i , and hence G = G for all subscripts s > r . I n this situation, there are just finitely m a n y groups appearing i n the tower, a n d G is the largest of these. In other words, T h e o r e m 9.10 is equivalent to the assertion that a complete group appears somewhere i n the a u t o m o r p h i s m tower of a n a r b i t r a r y finite group w i t h t r i v i a l center. O u r strategy of proof, however, w i l l not be to show directly t h a t some group i n the a u t o m o r p h i s m tower is complete. W e w i l l show instead t h a t if Z ( G ) = 1, then there is some integer n , depending o n l y on G , a n d such that | G * | < n for a l l terms G i n the a u t o m o r p h i s m tower of G . T h i s , of course, w i l l show that there are o n l y finitely m a n y different groups d occurring, as required.
r r r r r r s r %

It is not completely t r i v i a l to find examples where Z ( G ) = 1 and A u t ( G ) is not complete. B u t if there were no such example, then a u t o m o r p h i s m towers w o u l d never contain more t h a n two groups, a n d W i e l a n d t ' s theorem w o u l d not be very interesting. Before we begin w o r k i n g t o w a r d a proof of the a u t o m o r p h i s m tower theorem, therefore, it seems appropriate to m e n t i o n an example where there are three different groups i n the a u t o m o r p h i s m tower. Let G = S xS , the direct product of two copies of the s y m m e t r i c group of degree 3, and observe that Z ( G ) = 1. T h e n | G | = 36, and one can compute t h a t | A u t ( G ) | = 72 and | A u t ( A u t ( G ) ) | = 144. T h i s group of order 144, is complete; it can be constructed as the semidirect p r o d u c t of an elementary abelian group E of order 9, acted on by a semidihedral group of order 16, w h i c h is a full Sylow 2-subgroup of A u t ( ) ^ G L ( 2 , 3 ) .
3 3

P a r t s (a), (b) and (d) of the following establish the basic results about automorphisms to w h i c h we referred. 9.11. L e m m a . L e t A = A u t ( G ) a n d I = Inn(G), w h e r e G i s a n a r b i t r a r y g r o u p , a n d f o r g G G , w r i t e r t o d e n o t e t h e i n n e r a u t o m o r p h i s m of G i n d u c e d by g . T h e f o l l o w i n g t h e n h o l d .
g

(a) T h e m a p g ^ Z(G). (b) I < A .

is a homomorphism from G onto I , with kernel

(c) / / Z ( G ) = 1, t h e n C ( I ) = l .
A

(d) 7 / Z ( G ) = I , t h e n Z { A ) = 1. P r o o f . Since (x) = x ^ ) for elements x , g , h G , it is immediate that the m a p g i-> is a h o m o m o r p h i s m from G onto I . N o w r = 1 i f a n d
T g 9 h h

280

9. M o r e o n

Subnormality

o n l y if X = x for a l l x G G , and this is true if a n d only if g G Z ( G ) . It follows t h a t the kernel of the h o m o m o r p h i s m g ^ r is exactly Z ( G ) , and this proves (a).
9 g

N o w let a G A and g G G . T h e n for a l l x G G , we have (aO(f )

s 9 a

- ( a O a T ^ a = (((x)Q" ))a = ( ( x ^ a ) ^ "

FL Q

= ( x ) ^

and so ( T ) = T ( ) , w h i c h lies i n I . It follows that I < A , p r o v i n g (b).

Q

A s s u m e now that Z ( G ) = 1, and suppose t h a t a G C calculation. If g G G , we have

T

( I ) i n the previous

9 =

( <?) =

(g)a,

and since the m a p g i-> r is injective i n this case, it follows that # = (#)a. T h i s holds for a l l g G G , a n d so a = 1. T h i s establishes (c), and (d) is an i m m e d i a t e consequence.
5

Suppose that Z ( G ) = 1, and let G = G i < G < G < be the aut o m o r p h i s m tower of G . B y L e m m a 9.11(d), each of the groups G has a t r i v i a l center, a n d thus by 9.11(c), each of t h e m has a t r i v i a l centralizer i n its successor. T h e following result shows that i n this situation, C (G) = 1 for a l l z .
2 3 { Gi

It is convenient now to change notation. W e write S i n place of G , so that we can use the s y m b o l " G " to denote the largest group under consideration. 9.12. L e m m a . L e t S = Si < S w h e r e G i s a g r o u p , a n d assume T h e n C (S) = 1.
G 2

< S

< < S
S l + 1 t

= G, l < i < r .

that C

{ S ) = 1 f o r alli with

P r o o f . T h e result is t r i v i a l if r < 2, and so we assume that r > 2, and we proceed by i n d u c t i o n on r . W r i t e C = C { S ) , and observe that G n S _ i = 1 by the inductive hypothesis a p p l i e d i n the group S _ i . N o w S normalizes S, and so S normalizes G , and we have [ S , C ] C C. A l s o S C S _ i < G , and thus [ S , C ] C [ S _ i , G ] C S _ i . T h e n [ S , C ] C G n S -i = 1, and so C centralizes S . B y the inductive hypothesis i n G , w i t h S i n place of S and r - 1 i n place of r , we conclude that G C C ( S 2 ) = 1.
G r r 2 2 2 2 r 2 r r 2 r 2 2 G

N o w let G have t r i v i a l center, and suppose that its a u t o m o r p h i s m tower is G = G i < G < . R e c a l l that our goal is to show that there exists an integer n , depending only on G , and such that |G*| < n for a l l i. B y L e m m a 9.12, we k n o w t h a t C { G ) = 1, and thus the a u t o m o r p h i s m tower theorem w i l l be a n i m m e d i a t e consequence of the following result.
2 G i

9B

281

9 . 1 3 . T h e o r e m . L e t S G, w h e r e G i s a finite g r o u p , a n d assume that C (S) = 1. T h e n t h e r e exists a n integer n depending only o n the isomorp h i s m t y p e ofS, a n d such t h a t \ G \ < n .
G

O u r proof relies on two applications of the following easy lemma. 9 . 1 4 . L e m m a . L e t N be a n o r m a l s u b g r o u p of a finite g r o u p G, a n d suppose that N D C ( N ) . T h e n \ G \ d i v i d e s |Z(7V)||Aut(7V)|. I n p a r t i c u l a r , \ G \ divides \N\\.
G

P r o o f . T h e conjugation action of G on i V defines an isomorphism from G / C ( N ) into Aut(TV), and since C ( N ) = Z(7V), it follows that | G | d i vides | Z ( J V ) | | A u t ( 7 V ) | , as claimed. N o w A u t ( i V ) acts faithfully on the set of nonidentity elements of AT, and so |Aut(AT)| divides (|7V| - 1)!. Since | Z ( A ) | divides |AT|, the final assertion follows.
G G

R e c a l l t h a t i n an a r b i t r a r y finite group S, there is a unique n o r m a l subgroup m i n i m a l w i t h the property that the corresponding factor group is nilpotent. T h i s subgroup, denoted S , is the final t e r m of the lower central series of S. W i t h this notation established, and before we plunge into the details, we can give a brief outline of the proof of T h e o r e m 9.13 W e show first that if S is an a r b i t r a r y s u b n o r m a l subgroup of G , then the generalized F i t t i n g subgroup F * ( G ) normalizes S. If we can show t h a t N ( S ) has b o u n d e d order, this w i l l y i e l d an upper b o u n d on | F * ( G ) | , a n d since F * ( G ) contains its centralizer i n G, the result w i l l follow by L e m m a 9.14. I n order to o b t a i n an upper b o u n d on | N ( S ) | , we use the hypothesis that C { S ) = 1 to show t h a t C (S) C S. ( T h i s result of S c h e n k m a n seems to be the hardest step.) T h e n L e m m a 9.14 yields | N ( 5 ) | < | Z ( S ) | | A u t ( S ) | , a n d so this is an upper b o u n d for | F * ( G ) | , as needed.
G G G G 0O G 1 1

We begin work now w i t h a number of p r e l i m i n a r y results. 9 . 1 5 . L e m m a . L e t G be a finite g r o u p , a n d suppose S G a n d F < G, a n d assume that F is nilpotent. t h a t G = SF, w h e r e T h e n G = S.

P r o o f . Since there is n o t h i n g to prove if S = G , we can assume t h a t S < G , and we proceed by i n d u c t i o n on | G | . L e t S C M < G w i t h M < G. B y D e d e k i n d ' s lemma, M = S ( F n M ) , a n d so by the inductive hypothesis applied to M , w i t h F n M i n place of F , we have M = S. It thus suffices to show t h a t M = G . N o w M< G, a n d w e w r i t e G = G / M . Observe that G = M F since M contains S, and thus G = M F is a product of nilpotent n o r m a l subgroups, so it is contained i n F ( G ) , and hence is nilpotent. It follows that G / M is

282

9. M o r e o n S u b n o r m a l i t y

nilpotent, and so G C M . T h e reverse containment is clear since is nilpotent, and thus M = G . T h i s completes the proof. 9.16.
G

M/G

Then F ( G ) C

Corollary. L e t S

G, w h e r e G i s a

finite

group.

N (S). P r o o f . B y L e m m a 9.15, we have S = ( F ( G ) S ) , and since this subgroup is characteristic i n F ( G ) S , it is certainly n o r m a l i z e d by F ( G ) . We must also show that the layer E ( G ) normalizes S for every subn o r m a l subgroup S of G. Together w i t h C o r o l l a r y 9.16, this w i l l show that F * ( G ) normalizes S, as wanted. 9.17. L e m m a . L e t S G , w h e r e G i s a finite g r o u p a n d S i s n o n a b e l i a n and simple. Then the n o r m a l closure S i s a m i n i m a l n o r m a l s u b g r o u p of G, a n d s o S C S o c ( G ) .
G

P r o o f . R e c a l l that S is the subgroup generated by the members of the set X = { S | g G G } . Since the conjugates of S are simple, they are certainly quasisimple, a n d since they are s u b n o r m a l , they are components. It follows by T h e o r e m 9.4 t h a t the members of X normalize each other, and so S = n X and S is semisimple. T o show that S is m i n i m a l n o r m a l i n G , suppose t h a t 1 < N C S w i t h N < G. T h e n N contains some m i n i m a l n o r m a l subgroup of S . B y L e m m a 9.5, this m i n i m a l n o r m a l subgroup is a member of X , and thus N contains some conjugate of S. Since N is n o r m a l in G , it contains a l l conjugates of S, and it follows t h a t N = S . T h i s shows t h a t S is m i n i m a l n o r m a l i n G , as required.
9 G G G G G G G

9.18.

Corollary. L e t S S.

G, w h e r e G i s a

finite

group.

Then E(G) nor-

malizes

P r o o f . L e t Z = Z ( E ( G ) ) , and write G = G/Z. T h e n E ( G ) is semisimple, and by L e m m a 9.17, each of its simple factors is contained i n S o c ( G ) . T h e n E ( G ) C S o c ( G ) , and we conclude by T h e o r e m 2.6 t h a t E ( G ) normalizes the s u b n o r m a l subgroup S = ~SZ. It follows t h a t E ( G ) normalizes SZ, and so it normalizes ( S Z ) = S, where the equality follows by L e m m a 9.15.
0 0

O u r next goal w i l l be to show that if S G , and we assume i n a d d i t i o n t h a t C { S ) = 1, then C (S) C S. T h i s is a result of Schenkman, and we begin w i t h the following easy fact, w h i c h can also be obtained as a consequence of P r o b l e m I D . 1 5 .
G G

9.19. L e m m a . L e t N < G, w h e r e G i s a finite g r o u p , a n d suppose that N C $(G), the Frattini subgroup. I f G / N i s nilpotent, then G i s nilpotent.

9B

283

P r o o f . B y T h e o r e m 1.26, it suffices to show t h a t M < G, where M is an a r b i t r a r y m a x i m a l subgroup of G . B u t M Z} <&(G) Z2 N , so by the correspondence theorem, M / N is m a x i m a l i n the nilpotent group G/N. It follows t h a t M / N < G/N, and thus M < G. 9.20. L e m m a . L e t N be a n o r m a l s u b g r o u p of a finite g r o u p G, a n d assume t h a t G / N i s n i l p o t e n t . T h e n t h e r e exists a nilpotent subgroup H C G such t h a t N H = G. P r o o f . Since obviously, N G = G, the set of subgroups H C G such t h a t N H = G is nonempty, and we show that a m i n i m a l member of this set must be nilpotent. W e argue t h a t if H is m i n i m a l such that N H = G, t h e n N n H C $ ( i f ) . Otherwise, there exists a m a x i m a l subgroup M of H such that i V n H % M , and thus since i V n H < H , we see t h a t ( N n i f ) M is a subgroup of 7f t h a t properly contains M . T h e n (AT n i f ) M = 7f, and we have G = N H = N ( N C ) H ) M = N M . T h i s contradicts the m i n i m a l i t y of H , and thus N n H <Z $ ( i f ) , as c l a i m e d . N o w H / { N n i f ) = i V i f / i V = G / i V , and this is a nilpotent group. It follows t h a t by L e m m a 9.19 that i f is nilpotent. 9.21. C (S)
G

T h e o r e m (Schenkman). L e t S = 1. T h e n C ( S ) C S.
G

G, w h e r e G i s a

finite

group a n d

F i r s t , we prove the case of Schenkman's theorem where S = G. 9.22. T h e o r e m . C (G) C

G

L e t G be a

finite

g r o u p , a n d suppose

t h a t Z { G ) = 1.

Then

P r o o f . L e t C = C (G), and note t h a t C < G. W e argue that if i f C G is a nilpotent subgroup such t h a t GH = G, then C n H = 1. Otherwise, G n i f is a n o n t r i v i a l n o r m a l subgroup of the nilpotent group i f , and so it contains a n o n t r i v i a l central subgroup Z of H . T h e n H C C ( Z ) , and since Z C C , also G C C ( Z ) . T h e n G = GH C C ( Z ) , and this is a contradiction, since Z ( G ) = 1 by hypothesis.
G G G G

We w i l l show t h a t it is possible to choose a nilpotent subgroup H C G such that G i f = G and so that i n a d d i t i o n , G = ( G n G ) ( G n i f ) . Since G n i f = 1 by the result of the previous paragraph, it w i l l follow t h a t C = C n G , and this w i l l complete the proof. B y L e m m a 9.20, there exists a nilpotent subgroup K C G such t h a t G A : = G . W r i t i n g T = C K , we argue t h a t T C C n G . T o see this, observe first that each t e r m of the lower central series of T is contained i n the corresponding t e r m of the lower central series of G , and thus T C G . A l s o , T / G = C K / C is nilpotent because AT is nilpotent, a n d thus T C G , and this shows t h a t T C C D G , as wanted. B y L e m m a 9.20 applied

284

9. M o r e o n S u b n o r m a h t y

in the group T, therefore, we c a n find a nilpotent subgroup H such t h a t T = ( G n G)H. Since C n G C C C T , D e d e k i n d ' s l e m m a yields c = ( c n G ) ( C n i / ) . Also,
G P = G ( G n G)/f = G T D G A : = G ,

and thus G t f = G , so H has the desired properties. T h i s completes the proof.

P r o o f o f T h e o r e m 9.21. L e t G = C ( S ) , a n d observe t h a t C D S C S
G

by T h e o r e m 9.22. Since S normalizes S, i t also normalizes C, a n d hence SC is a group, a n d it is no loss t o assume t h a t SC = G. W e proceed b y Suppose that S C H < G. T h e n S tf, a n d since C (S)
H

induction on | G | . = 1 and H < G, the i n d u c t i v e hypothesis yields P n G

0 0

C S . B y Dedekind's lemma, G , we conclude T h e n G/S is

however, P = S ( H n G ) C S S

= 5 , a n d so H = S. It follows that there

are no subgroups P w i t h S < H < G, a n d since 5 cyclic, a n d since C / { C n S ) = G / ( G f l S ) = SC/S C/(CnS) Now
G

that S < G , a n d G / 5 has no nonidentity proper subgroups. is cyclic. B u t G n S C Z ( G ) ,

= G/S, we see t h a t

a n d we deduce t h a t G is abelian.

T h e n G = ( S G ) = S b y L e m m a 9.15.

Z ( G ) C C ( 5 ) = 1, so we c a n a p p l y T h e o r e m 9.22 t o G t o deduce
G

t h a t C ( G ) C G . Since G = S, the result follows.

We c a n now prove T h e o r e m 9.13, thereby c o m p l e t i n g the proof of W i e landt's a u t o m o r p h i s m tower theorem. P r o o f o f T h e o r e m 9.13. Since S < N ( S ) a n d T h e o r e m 9.21 guaranG

tees that S contains its centralizer i n N ( 5 ) , it follows b y L e m m a 9.14

G

that |N (5)| <

G

WS^WAutiS )].
G

00

N o w F * ( G ) = F ( G ) E ( G ) , a n d each factor is contained i n N ( S ) b y C o r o l laries 9.16 a n d 9.18, a n d thus | F * ( G ) | < | Z ( S ) | | A u t ( S ) | . Since F * ( G )

0O 0O

contains its centralizer i n G b y T h e o r e m 9.8, L e m m a 9.14 yields

|G| < |F*(G)|! < (IZ^
0 0

)!^^^

0 0

)!)!,

a n d thus | G | is bounded i n terms of S, as required.

Problems 9B 9 B . 1 . L e t S < G, where S is complete. subgroup T. Show t h a t G = S x T for some

9C

285

9 B . 2 . Suppose that Z ( G ) = 1, and let G = G i < G < be the automorp h i s m tower of G. If G < G show that i < 2.
2 it

9 B . 3 . If G is semisimple, show that A u t ( G ) is complete, a n d thus the aut o m o r p h i s m tower for G contains at most two different groups. 9 B . 4 . L e t G be dihedral of order 2 n , where n is o d d . Observe that Z ( G ) = 1, and show that the a u t o m o r p h i s m tower of G contains at most two different groups. 9 B . 5 . L e t G = A B , where A , B G . Show t h a t G = A ^ B
0 0

H i n t . F i r s t do the case where b o t h A and B are n o r m a l , and then work by induction on | G | .

9C We mentioned the Sims conjecture i n C h a p t e r 8. R e s t a t e d i n purely grouptheoretic language, this conjecture asserts t h a t there exists some function /(TO), defined for positive integers TO, such t h a t the following holds. L e t H C G be a m a x i m a l subgroup, and assume t h a t c o r e { H ) = 1. (In this situation, the subgroup H is said to be corefree.) L e t g e G w i t h g $ H , and letTO= \ H : H (~l H \ . T h e n \ H \ < /(TO).
G 9

It is a t r i v i a l i t y t h a t i fTO= 1 then \ H \ = 1, and ifTO= 2, then \ H \ = 2. If TO = 3, then \ H \ is a divisor of 48 by a 1967 theorem of C . Sims, but no general bounds for integers TO exceeding 3 were k n o w n to exist u n t i l 1983, w h e n P . C a m e r o n , C . Praeger, J . S a x l and G . Seitz proved the full conjecture in a paper relying on the classification of simple groups. T h e r e is, however, an elementary argument t h a t seems to come close to p r o v i n g the Sims conjecture, and w h i c h is the s t a r t i n g point of the paper by C a m e r o n et a l . T h i s result, proved by T h o m p s o n and simplified by W i e l a n d t , is an a p p l i c a t i o n of s u b n o r m a l i t y theory, and it is the subject of this section. T h o m p s o n proved that there is a function t { m ) such that i n the context of the Sims conjecture, \ H : O ( i f ) | < t ( m ) for some p r i m e p . I n other words, a l t h o u g h T h o m p s o n d i d not show t h a t \ H \ is bounded, he d i d prove t h a t if H is u n b o u n d e d l y large, t h e n it must be "mostly" a p-group.
p

9.23. T h e o r e m ( T h o m p s o n ) . L e t H be a c o r e f r e e m a x i m a l s u b g r o u p finite g r o u p G. L e t g e G - H , a n d w r i t e m \ H : H n H \ . T h e n f o r 2 p r i m e p , w e h a v e \ H : O { H ) \ < ((TO!) )!.

9 p

of a some

We prove the following somewhat more general form of this theorem. (The d i a g r a m should help the reader to keep track of the m a n y subgroups involved.)

286

9. M o r e o n S u b n o r m a l i t y

9.24. T h e o r e m . L e t H a n d K be d i s t i n c t s u b g r o u p s of a f i n i t e g r o u p G. W r i t e D H n K , a n d assume t h a t no n o n i d e n t i t y s u b g r o u p of D i s n o r m a l i n any s u b g r o u p of G t h a t p r o p e r l y c o n t a i n s e i t h e r H o r K . L e t M = c o v e { D ) a n d N = c o v e ( D ) , a n d l e t E = M C \ N . Then a t l e a s t one of t h e s u b g r o u p s U = c o r e { E ) a n d V = c o r e { E ) i s a p - g r o u p f o r some p r i m e p .

H K H K

K H

U n d e r the hypotheses of T h e o r e m 9.24, let a = \ H : D \ and b = \ K : D \ . B y the n!-theorem, we have \ H : M | < a! and \ K : N \ < b\. It follows t h a t \ D : E \ < \ D : M \ \ D : N \ < { a - 1)!(6 - 1)!, and thus \ H : E \ < a \ b \ and \ K : E \ < a \ b \ . B y the n!-theorem once again, \ H : U\ < { a \ b \ ) \ and \ K : V\ < ( a \ b \ ) \ . T h e o r e m 9.24 asserts t h a t one of U or V is a p-group for some prime p , and since U < H and V < K , we deduce that either \ H : O { H ) \ < { a \ b \ ) \ or \ K : O { K ) \ < ( a \ b \ ) \ .
P p

To see w h y T h e o r e m 9.23 follows from T h e o r e m 9.24, assume the hypotheses of 9.23, a n d suppose that H > 1. Since H is m a x i m a l and has t r i v i a l core, we have H = N ( H ) , and thus g does not normalize H , a n d we c a n take K = H , so that H and K are distinct. B y the m a x i m a l i t y of H again, the o n l y subgroup of G t h a t p r o p e r l y contains either H or K is G itself, a n d since H is corefree, no nonidentity subgroup of D = H n K is n o r m a l i n G. T h e hypotheses of T h e o r e m 9.24 are thus satisfied, and in the above n o t a t i o n , a = m = b. Since H and K are isomorphic, it
G 9

9C

287

does not matter w h i c h of U or V is a p-group since i n either case, we have \ H : O ( H ) \ < ((m!) )!, as asserted by T h e o r e m 9.23.
2 p

9.25. assume

L e m m a . L e t H C G, w h e r e G i s a finite that E(G) C H . Then E { G ) = E { H ) .

group and F(G) = I , and

P r o o f . Since E ( G ) C H , each component of G is contained i n H , a n d so the components of G are components of H , and we have E ( G ) C E { H ) . T o establish equality, it suffices to show that a l l components of H are contained in E ( G ) , so suppose that U is a component of H and U % E ( G ) . T h e n certainly U is not a component of G , and so each component of G is a component of H different from U. B y T h e o r e m 9.4, distinct components of H centralize each other, and it follows t h a t U centralizes a l l of the components of G , a n d thus U C C ( E ( G ) ) . B u t F ( G ) - 1, so F * ( G ) = E ( G ) , and thus U C C ( F * ( G ) ) C F * ( G ) = E ( G ) , a n d this is a c o n t r a d i c t i o n .
G G

N e x t , we prove a n analog of L e m m a 9.15. 9.26. L e m m a . L e t G = SP, w h e r e G i s a finite g r o u p , a n d w h e r e S G a n d P < G , a n d P i s a p - g r o u p f o r some p r i m e p . T h e n O P ( G ) = O ( S ) .

p

P r o o f . If S = G , there is n o t h i n g to prove, so we can assume S < G, and we choose M w i t h S C M < G and M < G. T h e n M = S { M n P ) , a n d so w o r k i n g by i n d u c t i o n on | G | and a p p l y i n g the i n d u c t i v e hypothesis to M , we have O { M ) = O ( S ) , and it suffices to show t h a t O ( M ) = O ( G ) .
p p p p

Now O { M ) < G , and^we wri_te_G = G / O ( M ) . Since S C M , we have G = M P , and thus G = M P , w h i c h is a p-group. It follows t h a t O ( G ) C O ( M ) . T h e reverse containment holds because M / O ( G ) is a p-group, and this completes the proof.
p p p p p

9.27. C o r o l l a r y . L e t G be a finite g r o u p . Suppose P < G be a p - g r o u p . T h e n P n o r m a l i z e s O ( S ) .

p

that S

G, and

let

P r o o f . B y L e m m a 9.26, we have 0 ^ ( 5 ) = O { S P ) . acteristic i n SP, so it is n o r m a l i z e d by P . U P r o o f of T h e o r e m 9.24.

T h i s subgroup is char-

Since V < K , we have V < M < H , so V H ,

and similarly, U K . A l s o , we can c e r t a i n l y assume that H > 1, and since F * ( H ) 2 C ( F * ( / / ) ) , it follows that F * ( P ) > 1, and thus either F ( H ) > 1 or E ( i f ) > 1, and similarly, we can assume t h a t either F(AT) > 1 or E(AT) > 1.
G

Suppose first t h a t F ( H ) = 1 a n d F(AT) = 1, so that E ( H ) > 1 and E(AT) > 1. W e show i n this case t h a t either U = 1 or V = 1, and so U or V is a p-group, as required. A s s u m i n g t h a t U > 1 and V > 1, we see t h a t neither U nor V is nilpotent, and thus U and V are n o n t r i v i a l subgroups of D ,

288

9. M o r e o n S u b n o r m a l i t y

n o r m a l i n H and AT, respectively. T h e n H = N ( P ) and AT = N ( V ) because n o n t r i v i a l subgroups of D cannot be n o r m a l i n subgroups properly c o n t a i n i n g H or AT. Since V P , it follows b y C o r o l l a r y 9.18 t h a t E ( P ) C N ( V ) = AT. T h e n E ( P ) C P>, a n d L e m m a 9.25 yields E ( P ) = E(P>). S i m i l a r l y , E(AT) = E ( D ) , and thus E ( P ) = E ( K ) is a n o n t r i v i a l subgroup of D , n o r m a l i n b o t h H and AT. T h e n i f = N ( E ( P ) ) = N ( E ( A T ) ) = AT, w h i c h is a c o n t r a d i c t i o n .
G G r G G G

We can now assume that either F ( H ) > metry, we can suppose that F { H ) > 1. T h e n a n d we w r i t e P = O { H ) . L e t X = ( y ( U ) that it suffices to show that at least one of X X > 1 a n d Y > 1, we have H = N ( A ) and derive a contradiction.
p G

1 or F ( K ) > 1, a n d by s y m O ( P ) > 1 for some prime p , and Y = O ( V ) , and observe or Y is t r i v i a l . A s s u m i n g that AT = N ( Y ) , and we work to
p p G

Since V H and P < # , it follows by C o r o l l a r y 9.27 t h a t P normalizes Y = O P ( V ) , and so P C N ( Y ) = AT. T h e n P C P , and i n fact, P < D since P < H .
G

Now D,
G

let k E K . W e have P < P , so P < AT, and hence U

f c f c k

<N

= N .
f c

Since N < D , we have P (N (A)) have P ^

f c k

3 F>, and by C o r o l l a r y 9.27 applied i n the group . Then P C N (A

G f c _ 1

it follows that P normalizes O P ( P ) = X = H , a n d we deduce that P C D , and thus P C D

p k f c _ 1 1

C P . Since also P

C AT, we

for a l l k E K . T h e n P C c o r e ^ ( P ) = A ,
p p

and so P < A ' , and we have O ( H ) = P C O ( A T ) C O ( K ) , containment follows since N < AT.

where the last

Now 1 < P C O ( A T ) , and so we can interchange the roles of P and K i n the previous argument. W e deduce t h a t O { K ) C O ( P ) , and so equality holds. T h u s P is a subgroup of D n o r m a l i n b o t h P a n d K , and since P > 1, we have H = N ( P ) = K , w h i c h is a contradiction. T h i s completes the proof.
p p p G

Problems 9C 9C.1. Suppose that b o t h P a n d K are s u b n o r m a l i n G i n T h e o r e m 9.24. P r o v e that either P = 1 or V = 1. H i n t . Otherwise, O ( G ) > 1 for some prime p . Show t h a t Z ( O ( G ) ) C H by considering separately the cases where P is or is not a p-group.
p p

9C.2.

L e t G = A B , where A , B

G . Show t h a t O { G ) =

OP(A)OP(B).

H i n t . A s w i t h P r o b l e m 9 B . 5 , first handle the case where A and P are b o t h n o r m a l , and then proceed by i n d u c t i o n o n \ G \ .

9D

289

9C.3. L e t G = ( A , B ) , where A , B G, a n d suppose that \ A : A ' \ and \ B : B ' \ are coprime. P r o v e t h a t G A B b y assuming t h a t G is a m i n i m a l counterexample to this assertion and c a r r y i n g out the following steps. (a) L e t N be m i n i m a l n o r m a l i n G. Show t h a t A N B = G.

(b) Show that ( N n A ) ( N n B ) is n o r m a l i z e d by b o t h A and B . (c) Show that N f] A = 1 N f] B , and N has prime order, say p . (d) N o w assume that p does not d i v i d e \ B : B ' \ . p r o b l e m to show that Q P ( G ) = Q P { A ) B . Use the previous

(e) D e r i v e a c o n t r a d i c t i o n using the fact that A Q P ( G )

is a group.

9D In the previous three sections, we presented applications of s u b n o r m a l i t y to general group theory. W e close this chapter w i t h a pretty result of D . B a r t e l s t h a t lies w i t h i n s u b n o r m a l i t y theory itself. O u r goal is to prove an analog for s u b n o r m a l subgroups of the obvious fact t h a t i f X C G, then the n o r m a l closure of X i n G, w h i c h by definition is the unique smallest n o r m a l subgroup of G t h a t contains X , is exactly the subgroup generated by all of the conjugates of X i n G. F i r s t , we observe t h a t for an a r b i t r a r y subgroup X C G, there is a unique smallest s u b n o r m a l subgroup of G t h a t contains X . T h i s subgroup is the s u b n o r m a l closure of X i n G, a n d it exists because intersections of s u b n o r m a l subgroups are s u b n o r m a l , a n d so the intersection of all s u b n o r m a l subgroups containing X has the desired properties. B y analogy w i t h the n o r m a l closure X , the n o t a t i o n X " is sometimes used to denote the s u b n o r m a l closure of X i n G. Observe t h a t we always have X " C X , and X G i f a n d only i f X = X " . Since X is generated by a l l of the G-conjugates of X and X " C X , it is perhaps n a t u r a l to guess that the s u b n o r m a l closure X " is generated by some of the conjugates of X .
G G G G G G G G G

G i v e n X , Y C G , we say t h a t X a n d Y are strongly conjugate i n G i f X and Y are conjugate i n the group ( X , Y ) , or equivalently, X and Y are conjugate i n every subgroup of G that contains b o t h of t h e m . ( B u t note t h a t strong conjugacy is n o t usually an equivalence relation; it is reflexive a n d s y m m e t r i c , but i n general, it is not transitive.) W e saw i n P r o b l e m 2 A . 1 0 that X G if and only if X has no strong conjugates i n G distinct from itself, a n d this suggests that perhaps i n general, the s u b n o r m a l closure of X is the subgroup generated by a l l of the strong conjugates of X i n G . T h i s is, i n fact true, and it is easy to see t h a t P r o b l e m 2 A . 1 0 is an i m m e d i a t e consequence.

290

9. M o r e o n S u b n o r m a l i t y

9.28. T h e o r e m ( B a r t e l s ) . L e t X C G , w h e r e G i s a X "
G

finite

group.

Then

= ( Y \Y

is strongly conjugate to X ) .

We write X ^ to denote the subgroup generated by a l l of the strong conjugates of X i n G. T h u s B a r t e l s ' theorem asserts t h a t X ( ) = X " , and so once we establish the theorem, the n o t a t i o n X ^ becomes superfluous.
G G

We X ^ \
G

begin work by n o t i n g some elementary properties of the subgroup F i r s t , observe t h a t if X and Y are strongly conjugate i n G and g G,
9

then X

and Y

are strongly conjugate, and thus ( X ( )

= (X )

{ G )

The

following l e m m a establishes a few slightly less t r i v i a l properties. 9.29. L e m m a . L e t I C G , w h e r e G i s a (a) I f X C S G, t h e n A < ) C S. C X ^ \ W C X ^ . = X ^ , a n d i n p a r t i c u l a r , if K =

G G

finite

group.

(b) I f Y C X C G, t h e n Y ^ (c) I f X C K C G, t h e n x

(d) If X ^ C i f C G, t h e n X X ^ \ thenK = X ^ l

P r o o f . W e prove (a) by i n d u c t i o n on | G | . L e t X C S G, and observe t h a t since there is n o t h i n g to prove if S = G, we can assume t h a t S < G. T h e n there exists a subgroup M < G w i t h S C M < G, and a l l conjugates of X i n G are contained i n M . T h e strong conjugates of X i n G , therefore, are exactly the strong conjugates of X i n M , and thus X ( ) = X ^ C 5 , where the containment holds by the inductive hypothesis a p p l i e d i n the group M .
G

For (b), let Y C X , a n d suppose that Y and Z are strongly conjugate in G . T h e n 2 = 7 C 1 for some element g G (Y, Z ) C ( X , X ) , and thus X a n d X a r e strongly conjugate i n G , and C X^ \ Therefore Z = Y C X C X ( ) , and since Y ( ) is generated by a l l such subgroups Z , it follows that Y ( ) C X ( ) , as wanted.
9 9 G 9 9 G G G

Now let X C AT C G . T h e strong conjugates of A i n AT are exactly those strong conjugates of A i n G that are contained i n K . T h u s X ^ C X ( \ and this proves (c). If X < ) C AT, then AT contains a l l strong conjugates of X i n G , so A ^ ) = X ( \ proving (d).
G G G

The

following is somewhat less routine. finite g r o u p G, a n d w r i t e

9.30. L e m m a . L e t N be a n o r m a l s u b g r o u p of a G = G/N. Then X ( ) = X

G { G )

for all subgroups

X o f G .

P r o o f . It is clear that the canonical homomorphism _g i-> p_maps the set of conjugates of A i n G onto the set of conjugates of X i n G . W e w i l l show

9D

291

that the set of_strong conjugates of X i n G maps onto the set of strong conjugates of X i n G, and the result then follows. F i r s t , suppose t h a t Y is strongly conjugate-to X i n G. T h e n Y = X some element g G ( X , Y ) , a n d we_have_Y = ( X ) and g G ( X , Y ) = ( X , T h u s Y is strongly conjugate to X i n G , as required.
9 9

for Y).

Conversely, consider a strong con}ugate of X i n G . Since this subgroup is conjugate to X , it has the form Y for some (not uniquely determined) conjugate Y of X i n G. L e t }> be the set of a l l conjugates Y of X i n G such that Y is the given strong conjugate of X , and observe t h a t it suffices to show t h a t some member of y is strongly conjugate to X . Choose Y _ G y so that the subgroup ( X , Y ) has the smallest possible order. Since X and Y are strongly conjugate, they are conjugate i n ( X , Y ) = ( X , Y ) , and thus there is some element g G ( X , Y) such that Y = ( X ) . T h e n NX = (NX) = NY, and thus X G y . A l s o , b y the choice of Y , we have \ ( X , X ) \ > \ ( X , Y ) \ . Since 3 G ( X , Y ) , however, we see that X C ( X , Y ) , and thus ( X , X ^ ) C ( X , Y ) . W e deduce t h a t ( X , X ) = ( X , Y ) , and since g is an element of this subgroup, X is strongly conjugate to X . B u t X G X and so the proof is complete.
9 9 9 9 9 9 9 9 9

We

also need the following easy general fact. G a n d P G Syl (G), w h e r e G i s a

p

9.31. L e m m a . L e t S T h e n P n S G Syl (S).

p

finite

group.

P r o o f . If S = G, there is n o t h i n g to prove, so we assume that S < G, and we work by i n d u c t i o n on | G | . L e t S C M < G w i t h M < G, and observe that \ M : M n P \ = \ P M : P \ is coprime to p . T h e n M n P e S y l ( M ) , and thus S n P = S n ( M C ) P ) e S y l ( S ) b y the i n d u c t i v e hypothesis.
p p

P r o o f of T h e o r e m 9.28. B y L e m m a 9.29(a), we have X ^ C X " . W e must prove the reverse containment, a n d for this purpose, it suffices to show t h a t X ( ) 3 G. Suppose that this is false, and assume t h a t | G | is as s m a l l as possible for a counterexample. A l s o , assume that | X | is as s m a l l as possible among subgroups X C G such t h a t X < ) is not s u b n o r m a l . N o t e that X ( ) < G , and hence X < G . W e proceed i n a number of steps.
G

Step 1. Suppose that Y and Z are conjugates of X and t h a t y W for some subgroup H containing Y and Z . T h e n Y ^ = Z ^ \
G G H H

P r o o f . W r i t e K = Y^ \ and let U = Y^ \ so t h a t also U = Z ^ \ W e have U = Z M and K = Y ^ by L e m m a 9.29(d), and U = Y ^ C Y ^ = AT by 9.29(c). Since Y is conjugate to X , we see that K = Y ^ is conjugate to

292

9. M o r e o n S u b n o r m a l i t y

< G, a n d thus K < G. T h u s Z(*> we have Y C P = Z

( c / )

K b y the m i n i m a l i t y of G, and
{ K )

C Z

K ,

where the second containment follows from 9.29(c) because U C K . T h e n y(K) ^ { K ) 9.29 (a), and we have
Z B Y

y (G) = ^ = y W c z<*> c z
We have just proved that y( >
G

^,

where the last containment follows by yet another a p p l i c a t i o n of 9.29(c). C Z ^

G

\ and since the reverse containment

follows symmetrically, we have equality, as required. Step 2. X is a p-group for some p r i m e p. P r o o f . Otherwise, since X is certainly generated b y its Sylow subgroups, it follows that X = ( Y | y < X ) . N o w write U = ( Y | Y < X ) and observe t h a t since Y C y( ), we have X C P , and also, [7 C X ^ by 9.29(b). B y the m i n i m a l i t y of X , each of the subgroups Y ^ for y < X is s u b n o r m a l i n G, and thus [/ <^ G. B u t I C [ / , so we have X ^ C [ / b y 9.29(a). Since we established the reverse containment previously, we have X ^ = UG. T h i s is a contradiction, and we conclude that X is a p-group, as claimed.
( G ) G

Step 3. L e t Y C H < G, where Y is conjugate to X , and let P S y l ( 7 f ) .

p

T h e n there exists Z C P w i t h Z conjugate to X , such that Y ^ P r o o f . Since H < G, we have Y ^ for some element / i Y
Z ( i i )

= Z^ \

7f, and thus P

( i i ) h

nY

( i i )

G
h

Syl (y( ))
p

ii

by L e m m a 9.31. A s Y is a p-subgroup of Y

, we can w r i t e Y

C Pn

. W r i t i n g Z = Y ,we
{

have Z C P and = yW ,

(H)

Y )

{ H )

= (y^f

and

thus Y (

G )

= Z(

G )

by Step 1.

Step 4. L e t M be a m a x i m a l subgroup of G c o n t a i n i n g X , and let P

p p

S y l ( M ) . T h e n P G S y l ( G ) , and M is the unique m a x i m a l subgroup of G containing P. P r o o f . F o r subgroups H C G , w r i t e /C(P) to denote the set of a l l subgroups of the form Y^ \
G

where Y C H and Y is conjugate to A i n G . If h G P and

1

Y C H is conjugate to X , then ( Y ^ ) ) ' acts by conjugation on the set of this action.

= (Y )

/ l

( G )

, and it follows t h a t H

/C(P). A l s o , Z (P) is contained i n the kernel

Now G acts t r a n s i t i v e l y on J C ( G ) , and M stabilizes (setwise) the subset IC(M) <ZK(G). T h e full stabilizer i n G of J C { M ) is thus a subgroup of G

9D

293

t h a t contains the m a x i m a l subgroup M , so it is either M itself, or else it is all of G. Suppose first that G stabilizes / C ( M ) . Since G acts t r a n s i t i v e l y on ( G ) , it follows that J C { M ) = K ( G ) . thus K ( G ) K < = /C(P). B y Step 3, we also have J C { M ) = / C ( P ) , and N o w Z ( P ) acts t r i v i a l l y on / C ( P ) = I C ( G ) , a n d it so t h a t | G | < | G | . B y L e m m a 9.30 and the

follows t h a t the action_of G on K { G ) has a n o n t r i v i a l kernel P . O f course, G, and we w r i t e G = G/K, m i n i m a l i t y of G, we have

and so A T A T ^ G. A l s o , X ( ) < K X ^ because AT stabilizes X < ) , and thus G. T h i s is a c o n t r a d i c t i o n , a n d it follows that G does not stabilize / C ( M ) . T h e n M is the full stabilizer of J C ( M ) = / C ( P ) , and this set is stabilized by N ( P ) . T h u s N ( P ) M , and since P G S y l ( M ) , it follows easily that P must be a full Sylow p-subgroup of G, as required. A l s o if P N , where N is m a x i m a l i n G, then P G Sylp(7V), and similar reasoning shows that N is the full stabilizer of / C ( P ) . T h u s A ' = M , as wanted.
G G p

Step 5. X is contained i n a unique m a x i m a l subgroup of G. P r o o f . Otherwise, let M and N be distinct m a x i m a l subgroups such t h a t X M n N , and let X S G S y l ( M n A"). Choose M and N so that | 5 | is as large as possible. If S G S y l ( G ) , then Step 4 yields M = A , w h i c h is a c o n t r a d i c t i o n . A l s o , observe t h a t X S, so if S < G, then X G. T h e n X ^ = X b y 9.29(a), so A ( ) is s u b n o r m a l , w h i c h is a c o n t r a d i c t i o n . T h u s N ( 5 ) is proper, and so it is contained i n some m a x i m a l subgroup R of G.
1 p p G G

N o w let S P G S y l p ( M ) and S Q G S y l ( A ^ ) , and note t h a t P and Q are full Sylow p-subgroups of G b y Step 4. T h e n S < P and S < Q, and so P n P > S and P n Q > S, a n d thus Sylow p-subgroups of M n P and N d R are larger t h a n S. B y the choice of M and TV, it follows t h a t M = R = N , w h i c h is a c o n t r a d i c t i o n .
p

Step 6. W e have a contradiction. P r o o f . W e have A ^ ) X < G, and thus A ^ ) cannot be s u b n o r m a l i n A , and it follows by the m i n i m a l i t y of \ G \ t h a t X = G.
G G G G G

L e t M be the unique m a x i m a l subgroup of G t h a t contains X . T h e n A g M , so some conjugate Y of A is not contained i n M . If ( A , Y ) is contained i n some m a x i m a l subgroup N , then X N , and so A = M , a n d this is a c o n t r a d i c t i o n because Y g M . T h u s ( A , Y) = G and Y is strongly conjugate to X . T h e n G = (X, Y ) X ( ) < G, w h i c h is a c o n t r a d i c t i o n .
G G

294

9. M o r e o n S u b n o r m a l i t y

Problems

9D

9 D . 1 . Just as s u b n o r m a l closures exist, so too do s u b n o r m a l cores. P r o v e t h a t i f l C G , then there exists a unique largest subgroup of X that is s u b n o r m a l i n G. Show that this s u b n o r m a l core of X i n G is n o r m a l i n X. 9 D . 2 . L e t I C G . P r o v e that the intersection L of a l l strong conjugates of X i n G is n o r m a l i n X and s u b n o r m a l i n G. H i n t . Show t h a t C X .

9 D . 3 . P r o v e that the s u b n o r m a l core of X i n G is not necessarily the i n tersection of any collection of conjugates of X . H i n t . L e t G = S , the s y m m e t r i c group of degree 4, a n d take X to be cyclic of order 4.
4

9 D . 4 . A converse of L e m m a 9.31 asserts t h a t U S Q G , and for a l l primes p a n d a l l Sylow p-subgroups P of G, the intersection P n 5 i s a Sylow p subgroup of S, then S is s u b n o r m a l i n G. Suppose that S C G yields a counterexample to this assertion w i t h \S\ + \ G \ as s m a l l as possible. Show that G and S are nonabelian simple groups. In particular, this converse of 9.31 holds if S is solvable. H i n t . Consider a m i n i m a l n o r m a l subgroup of G. N o t e . T h i s converse to L e m m a 9.31 was conjectured b y O . K e g e l . It was eventually proved by P . K l e i d m a n using the classification of simple groups.

Chapter 10

More Transfer Theory

10A We begin w i t h a brief review of some m a t e r i a l from C h a p t e r 5. L e t P be a Sylow p-subgroup of a finite group G , and let v : G -* P / P ' be the transfer h o m o m o r p h i s m . W e proved that k e r ( v ) is the subgroup A P ( G ) , w h i c h , by definition, is the smallest n o r m a l subgroup of G such that the corresponding factor group is an abelian p-group. O n e goal of transfer theory is to find a relatively s m a l l subgroup N , where P C N C G , a n d such that G / A * ( G ) = N/AP{N). (In other words, we want TV to control p-transfer i n G . ) If this happens, it might allow us to prove t h a t G is nonsimple by s t u d y i n g the smaller group N and showing that N has a n o n t r i v i a l abelian p-factor group. T h e n A ( N ) < N , and thus by c o n t r o l of p-transfer, A ( G ) < G. A s s u m i n g that G does not have order p, therefore, it cannot be simple.
P P

A n a t u r a l candidate for TV = N ( P ) , but this does to control p-transfer if G is N (P) = P , and A { P ) has
G p G

control of p-transfer is the Sylow normalizer not always work. F o r example, N ( P ) fails the alternating group A s , w i t h p = 2. T h e r e index 4 i n P , but G is simple.
G G

If the Sylow subgroup P is abelian, we proved i n C h a p t e r 5 that N ( P ) a c t u a l l y does control p-transfer i n G . T h i s was done i n two steps: first, L e m m a 5.12 shows t h a t N = N ( P ) controls G-fusion i n P , w h i c h means t h a t whenever two elements of P are conjugate i n G , they are already conjugate i n N , and then by C o r o l l a r y 5.22, we k n o w t h a t whenever F C J V C G and TV controls G-fusion i n P , it also controls p-transfer i n G .
G

If the Sylow p-subgroup P is nonabelian, the Sylow normalizer usually does not control fusion i n P . Nevertheless, even i n this case, there are situations where N ( P ) can be proved to control p-transfer. T h e m a i n result
G

295

296

10. M o r e Transfer

Theory

of this section is a powerful theorem of this type due to T . Y o s h i d a . It strengthens earlier results of P . H a l l and H . W i e l a n d t , and it guarantees that the Sylow normalizer controls p-transfer if the Sylow subgroup P is not too far from being abelian. M o r e precisely, N ( P ) controls p-transfer if the nilpotence class of P is less t h a n the p r i m e p. (If p = 2, this corollary of Y o s h i d a ' s theorem tells us n o t h i n g new, although the theorem itself has content for a l l primes.) O u r proof was inspired by Y o s h i d a ' s original character-theoretic argument, but it is more elementary t h a n his, and it does not rely on character theory.
G

For a given prime p , the key to Y o s h i d a ' s theorem is the wreath product W = C l C , where C is the cyclic group of order p . T h e group W can be defined as a semidirect product A x U , where A is the direct product of p copies of C and U ^ C . O f course, to construct the semidirect product W, we need a specific action of U on A , and to describe it we write
p p p p p

A = ( a i )

x (a > x x (a ) ,
2 p u a i i + l

and U = ( u ) , and we let U act on A by setting ( ) = a for 1 < i < p , and ( a ) = a . Following the usual custom, we view A and U as subgroups of W = C I C .
u p r p p

Before stating Y o s h i d a ' s theorem, we make a few routine observations about the group W = C I C . F i r s t , we see that the subgroup A of W satisfies \ W : A \ = p and | A | = j p . A l s o , A is elementary abelian, w h i c h means that it is abelian and t h a t x = 1 for a l l elements x A . Furthermore, A is generated by the subset { a a , , a } , w h i c h consists of p elements t h a t are conjugate i n W, and whose product is not the identity.
p p p u 2 p

In general, if P is an a r b i t r a r y nonabelian group h a v i n g an abelian subgroup A of p r i m e index p, then Z ( P ) C A . (Otherwise A Z ( P ) > A and thus P = A Z { P ) is abelian.) In this situation, if a A - Z ( P ) , then C { a ) = A , and hence the conjugacy class of a i n P has size p . A p p l y i n g this to the group W, we conclude that the generating elements <n form a full conjugacy class of W.
P

10.1. T h e o r e m (Yoshida). L e t P S y l ( G ) a n d N = N ( P ) , w h e r e G i s a finite g r o u p . T h e n N c o n t r o l s p - t r a n s f e r i n G unless P has a h o m o m o r p h i c image isomorphic to C \ C .

p G p p

10.2. m G .

Corollary. L e t P Syl (G) a n d N = N ( P ) , w h e r e G i s a

p G

finite

g r o u p . If t h e n i l p o t e n c e class

of P i s less

than p , then N controls p-transfer

P r o o f . B y L e m m a 10.3, below, the group C \ C has nilpotence class p, so it cannot be a h o m o m o r p h i c image of a p-group w i t h class less t h a n p. T h e result is now immediate from T h e o r e m 10.1.
p p

10A

297

O f course, we can also conclude t h a t N ( P ) controls p-transfer i n G if the Sylow p-subgroup P satisfies one of the several other conditions on a p-group that are sufficient to guarantee t h a t C } C is not a h o m o m o r p h i c image. F o r example, if P has exponent p, then C I C cannot be a hom o m o r p h i c image because C I C contains a n element of order p , namely u a i . M o r e generally, if for every two elements x,y P , there exists an element c i n the derived subgroup of ( x , y ) such t h a t { x y ) P = x P y P , then C i C cannot be a h o m o m o r p h i c image. ( A p-group that satisfies this cond i t i o n is said to be r e g u l a r . There is a fairly extensive theory of regular p-groups, w h i c h we w i l l not discuss here, but we m e n t i o n that the theorem of H a l l and W i e l a n d t that is generalized by Y o s h i d a ' s theorem asserts t h a t the normalizer of a regular Sylow p-subgroup controls p-transfer.)
G p p p p 2 p p p C p p

To a p p l y Y o s h i d a ' s theorem, one must show t h a t C I C is not a homom o r p h i c image of some given p-group, but to prove the theorem, we need conditions on a p-group sufficient to guarantee t h a t C I C a c t u a l l y is a h o m o m o r p h i c image. T h a t is the goal of the next few results.
p p p p

Since the group W = C l C satisfies the hypotheses of the following l e m m a w i t h t = p, the l e m m a implies t h a t \ Z ( W ) \ = p, t h a t W is elementary abelian of order p P ~ , and that the nilpotence class of W is p.
p p l

1 0 . 3 . L e m m a . L e t P be a p - g r o u p , a n d l e t A < P , w h e r e \ P : A \ = p , a n d A i s e l e m e n t a r y a b e l i a n of o r d e r p f o r some i n t e g e r t > 2. Suppose that A i s g e n e r a t e d by t h e elements of some c o n j u g a c y class of P . T h e f o l l o w i n g then hold.
l

(a)

\Z(P)\=p. a b e l i a n of o r d e r p " . of P i s t.
4 1

(b) P ' i s e l e m e n t a r y

(c) T h e n i l p o t e n c e class

P r o o f . L e t AT be a conjugacy class of P that generates A . Since \ A \ = p > p and A is elementary abelian, we see t h a t A is not cyclic, and thus the class K contains more t h a n one element. It follows t h a t P is nonabelian, and hence Z ( P ) C A . N o w P ' C A because P / A has order p, and so if a AT, we have ( a ) P ' C A . Since ( a ) P ' contains P ' , it is a n o r m a l subgroup of P , and hence it contains the entire conjugacy class K . B u t K generates A , and thus we have ( a ) P ' = A , and since \ {a) \ = p, it follows that \ A : P ' \ < p . T h i s yields p > | A | / | P ' | = | Z ( P ) | > p, where the equality follows .by L e m m a 4.6, and the final inequality holds since P is a p-group. T h u s | Z ( P ) | = p = \ A : P ' | , a n d (a) and (b) are proved. A l s o P has nilpotence class t by T h e o r e m 4.7, and thus (c) holds.

298

10. M o r e Transfer

Theory

B y w r i t i n g the group operation i n an elementary abelian p-group as add i t i o n , the group can be viewed as a vector space over the field Z / p Z of order p, and this enables us to use linear-algebra techniques. In particular, since the subgroups of an elementary abelian group are exactly the subspaces, and since every subspace of a finite dimensional vector space is a direct s u m m a n d , it follows t h a t i f X is an a r b i t r a r y subgroup of an element a r y abelian group E , then E = X x Y for some subgroup Y C E . T h i s observation and some other standard facts from linear algebra w i l l be used in the proof of the next result. R e c a l l t h a t the elements a of the group W = C I C , form a conjugacy class of W contained i n the elementary abelian subgroup A , and that the product of these elements is not the identity. Together w i t h the fact t h a t \ Z { W ) \ = p established i n L e m m a 10.3, this provides a useful way to recognize C I C , and this is the content of the following.
{ p p p p

1 0 . 4 . T h e o r e m . L e t P be a p - g r o u p , a n d l e t A < P , w h e r e A i s e l e m e n t a r y a b e l i a n a n d \ P : A \ = p . Suppose t h a t \ Z ( P ) \ = p , a n d assume that t h e p r o d u c t of t h e elements of some n o n c e n t r a l c o n j u g a c y class of P c o n t a i n e d i n A is not the identity. T h e n P ^ C l C .
p p

P r o o f . Since P has a noncentral conjugacy class, it is nonabelian, and thus Z ( P ) C A . B u t A is elementary abelian, and so we can write A = Z ( P ) x B , for some subgroup B C A . T h e n \ A : B \ = \ Z ( P ) \ = p , and we have \ P : B \ = p . A l s o , since B n Z ( P ) = 1 and every nonidentity n o r m a l subgroup of P meets Z ( P ) nontrivially, it follows t h a t c o r e ( P ) = 1, and thus the r i g h t - m u l t i p l i c a t i o n action of P on the p right cosets of B i n P is faithful. W e conclude t h a t P is isomorphic to a subgroup of the symmetric group S p 2 .
2 F 2

Let a G AT, where A" is a noncentral class of P such t h a t AT C A and the product of the elements of AT is not the identity. Choose u <E P - A , and l e t T : A ^ A b e the m a p x x . T h e n T is the identity m a p on A since u A and A is abelian. A l s o , AT = { a T | 0 < i < p } .
u p p l

Now view A as a vector space over the field F of order p . T h e n T is a linear operator on A and T = I , the identity operator on A . Since F has characteristic p , we have { T - I f = T - 1 = 0, and we let k be the smallest positive integer such that ( T - I ) = 0. T h e n k < p , and
p p k

A > A(T - I )> A(T - I)

> > A(T -

> A ( T - i f = 0,

a n d it follows that d i m ( A ) > k. N e x t , we compute w i t h p o l y n o m i a l s over the field F to o b t a i n (X - l )

1 1 p

- = ^

1 ) P

~ = 1 + X 4- X - l

X ~

10A

299

If fc < p , then 0 = (T - I f ' and thus a + a T + aT

2 1

= I + T + T

+ + T

+ + aT ~

= 0.

Since K = {aT* | 0 < i < p - 1 } , it follows t h a t (in the original m u l t i p l i c a t i v e notation) the product of the elements of K is the identity, and this is a contradiction. W e conclude t h a t k = p , and thus d i m ( A ) > p and | A | > p , and we have | P | > f .
p + l

B y the result of the first paragraph, P is isomorphic to a subgroup of the s y m m e t r i c group S 2 . Since the full p-part of { p ) \ is p P and we have shown that \ P \ > p P , it follows t h a t i n fact, P is isomorphic to a Sylow p subgroup of S 2 . B u t a l l Sylow p-subgroups of this group are isomorphic, and so P is uniquely determined up to i s o m o r p h i s m . T h e group C I C satisfies the hypotheses of the theorem, however, a n d it follows t h a t P C I C , as required.
2 + l p + 1 p p p p p

We mention t h a t the proof of T h e o r e m 10.4 yields another way to t h i n k about the group C \ C : it is a Sylow p-subgroup of the s y m m e t r i c group S 2 . W e shall not need this observation, however.
p p p

10.5. C o r o l l a r y . L e t P be a p - g r o u p , a n d l e t A < P , w h e r e A i s a n element a r y a b e l i a n s u b g r o u p of i n d e x p . L e t a A - Z ( P ) , a n d assume that t h e p r o d u c t of t h e p elements i n t h e c o n j u g a c y class of a i n P i s n o t t h e i d e n t i t y . T h e n C \ C i s a h o m o m o r p h i c i m a g e of P .

p p

P r o o f . L e t {a | 1 < i < p } be the conjugacy class of a, and write b = a i a - - - a p , so t h a t b 1 by assumption, a n d (b) C Z ( P ) . If | Z ( P ) | = p , then P ^ C I C by T h e o r e m 10.4, and there is n o t h i n g more to prove. W e can thus assume that Z ( P ) > ( b ) , and we let z Z ( P ) - ( b ) . T h e n z A , and thus Z = ( z ) has order p , and b Z .
2 p p

Now consider the group P = P / Z . Since b Z , we see that b is nontriv ial, a n d it is the product of the conjugate elements a l for 1 < i < p . If these factors were all equal, then since ( a i ) * = 1, we w o u l d have 6 = ( a ) P = 1, w h i c h is not the case.
x

Since { a l } is a conjugacy class of P contained i n A , and its c a r d i n a l i t y exceeds 1, it follows that this class has size p . T h e elements a~ are thus distinct, and the product of these elements is the nonidentity element b. T h u s P satisfies the hypotheses of the l e m m a , and w o r k i n g by m d u c t i o n on P , we can conclude that C I C is a h o m o m o r p h i c image of P . It is thus also a h o m o m o r p h i c image of P , as required.
p p

300

10. M o r e Transfer

Theory

W e also need to consider p-groups for w h i c h we are not given an element a r y abelian subgroup of index p . T h i s is where transfer becomes relevant, and so perhaps a brief review of some basic transfer theory w o u l d be useful. Let H C G have finite index, and let T be a right transversal for H i n G. (Recall t h a t this means t h a t exactly one member of T lies i n each right coset of H i n G.) If t G T and g G G, we w r i t e t>g to denote the unique element of T l y i n g i n the coset H t g . T h i s defines an action of G on T, and we see that tg[t-g)lies i n H . A pretransfer m a p V : G -* H is defined by the formula
1

V(g)

=
ter

Htg(t-g)-\

where the factors are m u l t i p l i e d according to some fixed, but a r b i t r a r y l i n ear order on the elements of T. In general, the m a p V depends on the transversal T and the p a r t i c u l a r order on the elements of T, but the comp o s i t i o n of V w i t h the canonical h o m o m o r p h i s m H H / H ' is the uniquely defined transfer h o m o m o r p h i s m v : G H / H ' , w h i c h is independent of the transversal and the linear order. In particular, since the transfer v is a hom o m o r p h i s m from G into the abelian group H / H ' , we see t h a t G' C ker(v), and thus if x , y G G and x = y m o d G', it follows that v ( x ) = v ( y ) , and thus V ( x ) = V { y ) mod H ' . A t one point i n our proof of Y o s h i d a ' s theorem, we w i l l need to consider the degenerate case where H = G. I n that s i t u a t i o n , we can take T = {1}, and we see t h a t the corresponding pretransfer is the i d e n t i t y map. It follows that i f V : G - > G is any pretransfer, then V { g ) = g m o d G'. 10.6. Lemma. Let M < P , where P is a p-group a n d \ P : M | = p , a n d

l e t V : P -> M be a p r e t r a n s f e r m a p . C h o o s e a t r a n s v e r s a l T f o r M i n P . Then f o r x G P , w e have (a) I f x e M , (b) I f x & M , t h e n V ( x ) = T\ x

ter
l

mod M ' .

then V ( x ) = x

mod M ' .

P r o o f . F i r s t , observe that m o d u l o M ' , the pretransfer V is uniquely determ i n e d , and so we can compute it using any convenient transversal. A l s o , since M < P , it follows that the set S = { r | t G T } is a transversal for M in P . (One way to see this is to observe that the coset M t ~ is the inverse of M t i n the factor group P / M , w h i c h is closed under t a k i n g inverses.)
1 l

U x e M , we use the transversal S to compute V ( x ) . W e have s x s " G M , and thus M s x = M s , a n d it follows t h a t s-x = s. T h e n s x ( s - x ) = sxs~ = x , where t G T and s = t~ . T h u s
1 l l l

V(x) =

JJ s x s -

JJ x*

mod M ' ,

10A

301

and

this proves (a).

i

Now let x P - M . T h e n the set P = { x | 0 < i < p ) is a transversal for M i n P , a n d this is the transversal we use to compute V ( x ) . T o determine u-x for u G P , consider first the case u = x for 0 < i < p - 1. C l e a r l y u-x = x , and the corresponding factor i n the c o m p u t a t i o n of V ( x ) is u x { u < x ) - = 1. T h e remaining case is u = x ^ " , where we have u-x = 1. The corresponding factor i n this case is u x ( u - x ) " = a*. M u l t i p l y i n g the factors for u U, we get V ( x ) = x? m o d M ' , as wanted.
{ i + 1 1 1 1

Before we state our next result, we recall t h a t if M is a p-group, then the F r a t t i n i factor group M / $ ( M ) is elementary abelian, a n d thus i n particular, M ' C $ ( M ) . I n fact, we can say more: $ ( A f ) / M ' is exactly the set of p t h powers of elements of the abelian group M / M ' . 10.7. Lemma. Let M < P , where P is a p-group and \ P : M | = p . L e t V : P ^ M be a p r e t r a n s f e r m a p , a n d assume that V ( M ) % $(M). Then C i C p i s a h o m o m o r p h i c i m a g e of P .
p

P r o o f . W r i t e P = P / * ( M ) , a n d observe t h a t since M ' C $ ( M ) , the element F ( x ) G M is uniquely determined for a l l elements x G P . B y h y p o t h esis, V ( M ) ^ $ ( M ) , and_thus we can choose a n element x G M such t h a t V(x) ^ $ ( M ) . T h e n V ( x ) + 1, a n d by L e m m a 10.6(a), we have

teT

where T is a transversal for M i n P . Since ( X ) P = 1 a n d _ ^ | = p , it follows t h a t the e l e m e n t e d are not a l l equal, a n d thus since M is elementary abelian a n d \ P : M | = p, we see that the elements x* are distinct, a n d they form a full conjugacy class i n P . T h e p r o d u c t of the elements i n this class is not the identity, and thus C o r o l l a r y 10.5 guarantees that C l C i s & h o m o m o r p h i c image of P . It is therefore also a h o m o m o r p h i c image of P , as wanted.
p p

N e x t , we present a general transfer theory result. 10.8. T h e o r e m ( T r a n s i t i v i t y of transfer). L e t G be a g r o u p , a n d t h a t H C K C G , w h e r e \ G : P | < oo. L e t U : G ^ K , W V be p r e t r a n s f e r maps. : K -> P : G - * H have V{g) = W ( U ( g ) ) mod H ' . and suppose

Then for all g e G , w e

302

10. M o r e Transfer

Theory

P r o o f . L e t T be the right transversal for K i n G that was used to construct the pretransfer m a p U. T h u s U(g) = I I
1 k

t.

where Jfe = t g ^ - g ) - A", and the factors A* are m u l t i p l i e d i n some definite but unspecified order. L e t w : K -> Tf/Tf' be the transfer, so t h a t is the coset of f f ' i n f f that contains W ( k ) for fc G K . Since ; is a h o m o m o r p h i s m , we have
t

w{U{g))
and thus

= Y[w(kt),
ter

W(U(g))

=l[W{kt)

mod f f ' .

ter N o w let 5 be the right transversal for f f i n K t h a t was used to construct the pretransfer m a p W. T h e n

w(k )=n
t
1 M

ht, ,
s

where / i = s k t ( s - k t ) G H , and where for each element t e T, the m u l t i p l i c a t i o n is carried out i n some definite but unspecified order. W o r k i n g m o d u l o H ' , the order of these factors is irrelevant, and we have

W(U(g))=l[h ,
t

mod/P.

We w i l l show t h a t the set ST = { s t | s e S, t G T } is a right transversal for H i n G, and that ( s i ) 5 ( ( s i ) ^ ) = h , . It w i l l follow t h a t V ( g ) is congruent m o d u l o H ' to the product (in any p a r t i c u l a r order) of the elements h , for s e S and t e T , and this w i l l complete the proof.
_ 1 t a t s

Let x G G be an a r b i t r a r y element. T h e n K x = K t for some element t G T , and we have x t ' G K . T h e n P ^ i " ) = P s for some element s G S, and thus / f x = H s t . T h i s shows that the coset H x contains st, and we must show that st is the unique element of ST contained i n H x . Suppose then t h a t H s t = H s ' t ' , where s' G S and t' G T. Since f f s and f f s ' are contained i n K , it follows t h a t K t = K t ' , and thus t = t! because T is a right transversal for K i n G. N o w f f s i = f f s Y = H s ' t , and hence f f s = H s ' , and we conclude t h a t s = s' because 5 is a right transversal for f f i n f i . T h i s shows t h a t ST is indeed a right transversal for f f i n G, as claimed.
1 1

W r i t e ( s t ) - g = s't', where s ' e S and ' G T . T h e n H s t g = f f s Y , and since f f s and f f s ' are contained i n K , we have K t g = K t ' , and thus t - g = t' and k = t g ( t - g ) - = t g ( t ) - . A l s o ,
1 1 t

f f sfc = H s t g { t ' )

= Hs't'{t')-

= Hs',

303

and thus s-k

= s' and h ,
t

= skt(s-kt)l

= skt(s')- . h
t : S

U s i n g the equations

k we o b t a i n

= tg{t'y

and

sk {s')-\
t

(stg^istyg)-

= = =

stg(s't')1

stg(t')- ( ')S

skt(s')t s

= h , Now m u l t i p l i c a t i o n over t G T and s G 5 yields Y(#) = VK(t/(#)) m o d i f ' , as wanted. F i n a l l y , we state a fairly technical sufficient c o n d i t i o n for a p-group to have a h o m o m o r p h i c image isomorphic to C \ C . It is this result t h a t we w i l l use i n the proof of Y o s h i d a ' s theorem. R e c a l l t h a t we write o { x ) to denote the order of an element x .
p p

10.9. T h e o r e m . L e t S < P , w h e r e P i s a p - g r o u p , a n d l e t V : P - > S be a p r e t r a n s f e r . F i x a n element x G P , a n d w r i t e R = (s e S \ o { s ) < o { x ) ) . If V(x) &R$(S), then C l C

p p

i s a h o m o m o r p h i c i m a g e of P .

P r o o f . Since S < P , we can choose a m a x i m a l subgroup M of P w i t h S C M , and we let U : P M and W : M -> S be pretransfers. If U { x ) 6 $(M), then f/(x) is a p t h power m o d u l o M ' , and we can write U ( x ) = y mod M ' for some element y M . W o r k i n g m o d S', the pretransfer W yields the transfer h o m o m o r p h i s m w : M - c 5 / 5 ' , a n d M ' C ker(w) since to maps to an abelian group. W e thus have
p

V(x) = W(U(x)) = W(y )

= W{y)

mod

S',

where the first congruence holds by T h e o r e m 10.8; the second holds because U ( x ) = y m o d M ' and M ' C ker(;), a n d the t h i r d congruence holds because w is a h o m o m o r p h i s m . B u t W ( y ) G S, so W { y ) G $ ( 5 ) , and thus V ( x ) G $ ( 5 ) since S" C $ ( 5 ) . T h i s is a c o n t r a d i c t i o n since we assumed t h a t V ( x ) # R $ ( S ) , and we conclude t h a t U ( x ) $ ( M ) . If x G M , the result follows by L e m m a 10.7.
p p

We can assume now that x & M , and thus x ^ 1 and o ( x ) < o ( x ) . W e work to derive a contradiction i n this case. L e m m a 10.6(b) yields x = U ( x ) mod M ' , and thus V ( x ) = W ( U ( x ) ) = W ( x ) m o d S'. (Note t h a t we cannot carry this one step farther and write W { x ) because W is defined only on M , but x M . )
p p p

B y the transfer-evaluation l e m m a ( L e m m a 5.5), we know t h a t W ( x ) is congruent m o d u l o S' to a product of conjugates of x l y i n g i n S. These conjugates are elements of S h a v i n g smaller order t h a n x , and thus they lie
p

304

10. M o r e Transfer

Theory

in R C i ? $ ( 5 ) . Since also S' C J * ( S ) , we conclude t h a t contrary to hypothesis.

i?$(S),

A s we mentioned, w h e n Y o s h i d a proved his result, he used character theory, and i n particular, he appealed to a representation-theoretic result k n o w n as M a c k e y ' s theorem. W e replace this w i t h the next result, w h i c h is a transfer analog of M a c k e y ' s theorem. T o state the M a c k e y transfer theorem, we consider the decomposition of a group G into ( H , AT )-double cosets, where H and K are a r b i t r a r y subgroups of G. These double cosets are the subsets of G of the form H g K , where g G G. Clearly, H g K is the orbit containing g of the action of the external direct product H x AT on G defined by g - ( h , k ) = h~ gk. Since the sets H g K are orbits i n an appropriate action, we see t h a t distinct ( H , AT)-double cosets are disjoint, and they p a r t i t i o n G.
l

We say t h a t a subset A C G is a set of representatives for the ( H , AOdouble cosets i n G if each of these double cosets contains exactly one element of X . If X is such a set of representatives, then the distinct ( H , AT)-double cosets i n G are exactly the double cosets H x K for x G A , and i n particular, | A | is equal to the number of these double cosets. N o t e t h a t the double coset H g K is invariant under left m u l t i p l i c a t i o n by elements of H , and thus it is a u n i o n of right cosets of H . Since G acts by right m u l t i p l i c a t i o n on the right cosets of H , the subgroup K acts by right m u l t i p l i c a t i o n on these cosets, and it is easy to see that the right cosets of H t h a t comprise the double coset H g K are exactly the members of the AT-orbit containing H g i n this action. Now assume that \ G : H \ < oo, so t h a t H has just finitely m a n y right cosets i n G. T h e n each double coset H g K is the disjoint u n i o n of the finitely m a n y right cosets of H t h a t it contains. These cosets of H form the AT-orbit containing H g , and so we can use the fundamental counting principle to determine their number. T h e stabilizer of H g i n G is H , and it follows that the size of the AT-orbit containing H g is |AT : K n H \ , a n d so this is the number of right cosets of H i n the double coset H g K . In particular, this number is finite, and so we have |AT : AT n H \ < oo. Furthermore, since the t o t a l number of right cosets of H i n G is finite and each ( H , AT)-double coset is a u n i o n of some of these, it follows from the fact that the distinct double cosets are disjoint that there are only finitely m a n y of them, and thus if X is a set of representatives for the ( H , AT)-double cosets i n G, we have | X \ < oo.
9 9 9

1 0 . 1 0 . T h e o r e m (Mackey transfer). L e t X be t h e ( H , K ) - d o u b l e cosets m a g r o u p G, w h e r e H \ G : H | < oo. L e t V : G - c H be a p r e t r a n s f e r x X , letW : AT -+ AT n H be a p r e t r a n s f e r

x x

a set of r e p r e s e n t a t i v e s f o r a n d AT a r e s u b g r o u p s , a n d m a p , a n d f o r each element m a p . T h e n f o r k G AT, w e

10 A

305

have V { k ) = JJ xex P r o o f . F i r s t , observe t h a t W { k ) K n P C H , and that W ^ f c ) is uniquely determined m o d u l o (AT n H ) ' C ( i i ) ' = ( P ' ) - It follows that xW^A^x G P , and this element is uniquely determined m o d u l o P ' . T h u s the p r o d u c t on the right is an element of H t h a t is uniquely determined m o d u l o H ' , and the order i n w h i c h the factors is m u l t i p l i e d is irrelevant.
x x x x 1 x - 1

x W x ^ x -

mod H ' .

For each element x X , let S be the right transversal for K n H i n K used to construct the pretransfer W . T h e n 1^1 = \ K : AT n P | , w h i c h is the number of right cosets of P i n H x K . I n fact, these right cosets form the AT-orbit of H x under right m u l t i p l i c a t i o n . Since A T n P is the stabilizer of H x i n AT, it follows that the elements of a right transversal for K D H in A" carry the coset P x to a l l of the different members of this P - o r b i t . T h i s shows that the cosets H x s for s G S are a l l of the right cosets of H i n H x K . A s x runs over A and s runs over S*, therefore, we account for each right coset of H i n G exactly once, and thus the disjoint u n i o n
x 9 x x X x

T =

\ J

xS

is a right transversal for H i n G. It follows t h a t V ( k ) E[] xex J] ^ ( ( a s ) . * : ) ses

x 1

mod H '

for a l l elements k G G . W e w i l l complete the proof by showing t h a t if k G K , then for a l l elements x G A , we have J] ses s s f c i i x s ) - * ; ) - = xWsCAOaT
1 1

mod H ' .

We need to compute ( x s ) - k , where x G X , s G and k G AT. T h i s element of T lies i n the coset H x s k , w h i c h is contained i n the double coset H x K because s G 5 C K a n d fc G K . It follows t h a t (xs)-fc lies i n xS , and so we can w r i t e ( x s ) - k = xs', where s' G 5 . T h e n H x s k = H x s ' , and if we left-multiply by x " \ we get ( H ) s k = ( H ) s ' , and so s k i s ' ) ' Tf*. B u t also s k i s ' ) ' G AT since s,s' S C K a n d jfc G K , and thus we have s k i s ' ) ' e K H H . It follows t h a t
X x X x x 1 1 x 1 x

iKC\H )sk and

= iKr\H )s',

thus s-k = s', where here, the dot denotes the a c t i o n of K on the right
x

transversal S

for K n i P i n AT. T h e n
1

xsfc((xs).fc)-

= xs^xsT

= xsfc^)" ^"

= x^s-A;)- )*- ,

306

1 0 . M o r e Transfer

Theory

and this is a n equality of elements of P . N o w m u l t i p l y over s G S fixed but a r b i t r a r y order, and work m o d u l o H ' to get J]
seS
x

i n some

xsk{{xs)-k)

=x(H
seS
x

skis-ky^x-

= xW {k)x~
x

mod H ' ,

as wanted.

Y o s h i d a ' s theorem asserts something about a Sylow p-subgroup P of G under the a s s u m p t i o n that N ( P ) does not control p-transfer. T h e following l e m m a w i l l allow us to exploit the failure of transfer control i n the proof of Y o s h i d a ' s theorem.
G

1 0 . 1 1 . L e m m a . L e t P C N C G, finite group G and N is a subgroup T h e n t h e r e exists a subgroup M < U { G ) C M f o r a l l p r e t r a n s f e r maps

w h e r e P i s a S y l o w p - s u b g r o u p of t h e t h a t does n o t c o n t r o l p - t r a n s f e r i n G. N , w i t h \ N : M \ = p , a n d such that U : G ^ N .

P r o o f . L e t V : G - > P and W : N -* P be pretransfer maps, and let v and w be the corresponding transfer homomorphisms o b t a i n e d by composing V and W w i t h the canonical h o m o m o r p h i s m P - > P / P ' . If U : G ^ N is a pretransfer, it follows by t r a n s i t i v i t y of transfer ( T h e o r e m 10.8) that v ( G ) = w ( U ( G ) ) C w ( N ) . Now G / A ? { G ) ^ v ( G ) and N / A P ( N ) = w ( N ) , and since we are assuming t h a t G / A P { G ) and N / A ( N ) are not isomorphic, it follows t h a t v ( G ) ^ w ( N ) , and thus v ( G ) < w ( N ) .
P

Since w ( N ) is a p-group and v ( G ) is a proper subgroup, there exists a subgroup L such that v ( G ) C L < w { N ) and \ w ( N ) : L \ = p . B y the correspondence theorem applied to the h o m o m o r p h i s m w : N P / P ' , the inverse image of L i n i V is a subgroup M of N such t h a t M < N , and \ N : M \ = \ w { N ) : L \ = p . A l s o , since w ( U ( G ) ) = v { G ) C L for an a r b i t r a r y pretransfer m a p U : G - > N , it follows that C/(G) C M , as wanted. F i n a l l y , we are ready to prove Y o s h i d a ' s theorem. P r o o f of T h e o r e m 1 0 . 1 . A s s u m e that AT = N ( P ) does not control ptransfer i n G , and let V : G -> A^ be a pretransfer map. B y L e m m a 10.11, we can choose a subgroup M < N w i t h |AT : M | = p, and such t h a t V ( G ) C M , and we note t h a t N ' C M .
G

Choose an element n G A" M such t h a t the order o(n) is as small as possible. If g is any p r i m e divisor of o ( n ) , then o ( n ) < o(n), and so n G M , and thus the image of n i n N / M has order g. B u t \ N / M \ = p, and it follows t h a t g = p, and thus o(n) is power of p. Since P is the unique Sylow p-subgroup of N , we have n G P .
9

Now let A be a set of representatives for the ( N , P ) - d o u b l e cosets i n G , and for each element x G A , fix a pretransfer m a p : P -> P n A/" .
s

P r o b l e m s 10 A

307

Since n G P , we can a p p l y the M a c k e y transfer theorem (Theorem 10.10) to deduce that

v

( ) = II xW {n)x~ xex
x

m o d TV'.

A l s o , V ( n ) G V"(G)

C M , a n d since also N ' C M , it follows that

Since P C N , one of the ( A , P ) - d o u b l e cosets is A/"IP = N , a n d so exactly one member of X lies i n N . If we call this element y , we see that P n A/T = P a n d thus Wj, is a pretransfer from P to itself, a n d we have W (n) = n m o d P ' . Since P ' C N ' C M a n d n G" M , it follows that W (n) g M = M , a n d thus y W { n ) y g M .
y y x y y

The product of the elements x W { n ) x lies i n M , a n d since the factor corresponding to x = y does not lie i n M , it follows that some other factor must also fail to lie i n M . T h e r e exists an element x G G - N , therefore, and a pretransfer m a p W : P -> P n N , such that x W ^ x ' M . We thus have W ( n ) 6 ( P n A ^ ) - ( P n M ) .
x x 1 x

Now w r i t i n g 5 = P C\ N and Q = P fl M , we have Q C 5 C P and VK(n) e S - Q. B y the choice of the element n , we know that a l l elements of N w i t h order smaller t h a n o ( n ) lie i n M , a n d thus a l l elements of N w i t h order smaller t h a n o(n) lie i n M . It follows that the subgroup R = (s G S | o(s) < o(n)) is contained i n Q . A l s o ,
x x x 1

\S : Q \ = \S : S n M and W(n)

= \SM

:M | < \ N : M

\ = p ,

therefore, $ ( 5 ) C Q . T h u s P $ ( 5 ) C Q, a n d since W ( n ) g Q , we have ^ i?*(5).

x x G x X

F i n a l l y , recall that x TV = N ( P ) , a n d thus P j i P . B u t P is the unique Sylow p-subgroup of G contained i n N , a n d thus P N , a n d we have S = P C \ N < P . B y T h e o r e m 10.9, therefore, C \C is a h o m o m o r p h i c image of P .
X V V

Problems

10A

1 0 A . 1 . Show that a 2-group P is regular if a n d only if it is abelian. H i n t . F i r s t , observe that factor groups of regular p-groups are regular. If P is a m i n i m a l counterexample to the p r o b l e m , show that | P ' | = 2.

308

10. M o r e Transfer

Theory

1 0 A . 2 . L e t P be a regular p-group. Show that the set { x e P \ x a subgroup.

= 1} is

H i n t . F i r s t , observe t h a t subgroups of regular p-groups are regular. Suppose t h a t P is a m i n i m a l counterexample to the p r o b l e m . Show that i f x,y P , where x has order p, then some proper subgroup of P contains b o t h x a n d x, a n d deduce that [ x , y f = 1. C o n c l u d e t h a t the p t h power of every element of P ' is t r i v i a l . 1 0 A . 3 . L e t P be a p - g r o u p w i t h | Z ( P ) | = p , a n d let A be abelian a n d have index p i n P . If Z ( P ) is a direct factor of A , deduce that P is isomorphic to a subgroup of C I C , a n d show t h a t A is elementary abelian.
p p

1 0 A . 4 . Suppose t h a t P e S y l ( G ) , a n d that P is isomorphic t o the direct

2

p r o d u c t of the quaternion group Q G' < G.

2 2 8

w i t h the cyclic group C . Show that

2

H i n t . Observe that C l C = D . A l s o , i f N = N ( P ) , show t h a t N / P has a n o n t r i v i a l fixed point i n its n a t u r a l action o n P / $ ( P ) .

G

1 0 A . 5 . L e t Q C P , where P is a p-group a n d | P : Q\ = p , a n d suppose that Q n Z ( P ) = 1. Show t h a t P is isomorphic t o a subgroup of C l C .

p p

1 0 A . 6 . L e t G act transitively o n a set ft, a n d let P be a point stabilizer. Show that G is 2-transitive o n ft i f a n d o n l y i f G = P U H g H for some element g e G. 1 0 A . 7 . L e t A < P , where \ P : A \ = p a n d A is a n elementary abelian p group. Suppose t h a t P - A contains a n element of order p a n d also a n element of order p . Show that C l C i s a h o m o m o r p h i c image of P .
2 p p

10B A group is metacyclic if it has a cyclic n o r m a l subgroup w i t h a cyclic factor group. I n this section, we use Y o s h i d a ' s theorem t o prove the following result of B . H u p p e r t . 10.12. T h e o r e m \G : G'|. We begin w i t h some easy observations. group. (Huppert). L e t P e S y l ( G ) , w h e r e G i s a
p

finite

group

a n d p > 2 , a n d suppose

that P is nonabelian and metacyclic.

T h e n p divides

10.13. L e m m a . L e t P be a m e t a c y c l i c (a) I f U o P , (b) I f V C P ,

then P / U is metacyclic. then V is metacyclic.

1 0 B

309

P r o o f _ L e t _ C < P,_where C a n d P / C are cyclic. F o r (a), w r i t e _ P = P / U . T h e n C < _ P , a n d C is a h o m o m o r p h i c image of C, a n d thus C is cyclic. A l s o , P / C ^ J P J U C , a n d this is a h o m o m o r p h i c image of the cyclic group P / C . T h u s P / C is cyclic, a n d hence P is m e t a c y c l i c , as required. F o r (b), observe t h a t V C \ C < V a n d V n C is cyclic since it is a subgroup of C. A l s o , n C ) ^ FC/C, w h i c h is a subgroup of P / C , a n d so it is cyclic. It follows t h a t V is m e t a c y c l i c .
I

10.14. L e m m a . Assume t h a t p > 2. T h e n t h e g r o u p W = C \ C h o m o m o r p h i c i m a g e of a m e t a c y c l i c g r o u p .

V

isnot a

P r o o f . B y L e m m a 10.13(a), it suffices to show t h a t W is not metacyclic. If W is metacyclic, however, then W is contained i n a cyclic n o r m a l subgroup, and so W is cyclic. T h i s is not the case, however, since W is elementary abelian of order p P ~ > p by L e m m a 10.3(b).
l

Now assume the hypotheses of T h e o r e m 10.12. A s we have just seen, C l C cannot be a h o m o m o r p h i c image of the m e t a c y c l i c Sylow p-subgroup P , a n d hence T h e o r e m 10.1 guarantees that N = N ( P ) controls p-transfer in G. If we c a n show t h a t A { N ) < N , therefore, it w i l l follow t h a t A ( G ) < G, a n d thus p divides \ G : G'\, as wanted. It suffices, therefore, t o prove the following result.
p p G P P

10.15. T h e o r e m . L e t P < N , w h e r e N i s a finite g r o u p a n d P e S y l ( A 0 , a n d assume t h a t P i s n o n a b e l i a n a n d m e t a c y c l i c , a n d t h a t p > 2. T h e n p divides \ N : N ' \ .

p

We need Maschke's theorem, w h i c h is usually viewed as belonging to representation theory, b u t w h i c h is useful w h e n considering coprime actions on elementary abelian p-groups. W e use the more-or-less standard proof of Maschke's theorem to establish T h e o r e m 10.16, whose statement is somewhat more general t h a n the usual statement of M a s c h k e ' s theorem. 10.16. T h e o r e m ( M a s c h k e ) . L e t K be a g r o u p of finite o r d e r m, a n d s u p pose t h a t K acts v i a a u t o m o r p h i s m s o n a g r o u p V w i t h s u b g r o u p s U a n d W such t h a t V = U xW, w h e r e U i s a b e l i a n a n d K - i n v a r i a n t . Assume that t h e m a p u ^ u i s b o t h i n f e c t i v e a n d s u r j e c t i v e o n U. T h e n t h e r e exists a K - i n v a r i a n t s u b g r o u p N < V such t h a t V = U x N .
m

Observe that i f U is finite a n d has order coprime t o m , then the m a p u i-> u is a u t o m a t i c a l l y b o t h injective a n d surjective. T h i s is because it has a n inverse, namely the m a p u i-> u , where the integer n is chosen so that m n = 1 m o d | P | .
m n

In the standard version of Maschke's theorem, V is a vector space over some field P , a n d U is a given AT-invariant subspace. I n t h a t case, it is

310

10. M o r e Transfer

Theory

a u t o m a t i c t h a t U is a direct s u m m a n d of V, a n d so no explicit mention of W appears i n the usual statement of the result. Here, of course, the group V is w r i t t e n additively, a n d so the key requirement is t h a t the m a p u ^ m-u should be injective a n d surjective, a n d this happens i f the characteristic of the field F does not d i v i d e TO. ( A n d it fails otherwise.) W e need the following case of Maschke's theorem for the proof of Theorem 10.15. 10.17. C o r o l l a r y . L e t K be a finite g r o u p t h a t acts v i a a u t o m o r p h i s m s o n a n e l e m e n t a r y a b e l i a n p - g r o u p V, a n d assume t h a t p does n o t d i v i d e \ K \ . Suppose t h a t U C V i s a K - i n v a r i a n t s u b g r o u p . T h e n t h e r e exists a K - i n v a r i a n t s u b g r o u p N C V such t h a t V = U x N . P r o o f . W r i t e TO = \ K \ a n d observe that \ U \ is coprime to TO, SO that as we observed, the m a p u ^ u is b o t h injective a n d surjective o n U, as required for T h e o r e m 10.16. It suffices, therefore, to show that V = U x W for some subgroup W C V. A s we mentioned i n the discussion preceding T h e o r e m 10.4, this follows easily by viewing V as a vector space over the field Z / p Z .
m

P r o o f of T h e o r e m 10.16. Since V = U x W, we c a n define the m a p 0 : V U b y setting 9 { u w ) = u i o r u e U a n d w e W ,

k

a n d we observe that 6 is
k k 1

a h o m o m o r p h i s m a n d t h a t 6 { u ) = u for u E U. N o w let x E V a n d k E K . T h e n 0 ( x ) G U, a n d since U is AT-invariant, we see t h a t 6 ( x ) ~ Since K is finite, we c a n define the m a p <p : V - > U b y E U.

^ ( x ) = J]
keK

e{x ) ~\
k k

a n d we observe t h a t this is well denned since each factor i n the product lies in U a n d U is abelian. I n fact, <p is a h o m o m o r p h i s m since i f x , y E V a n d k E K , w e have { x y ) = x y , a n d hence
k k k

rtxv)

n keK

^ ( ^ r

l[(e(x )e(y )) ~
k k k

= n e(xk)k~l n
keK k e K

where, of course, the t h i r d equality holds because U is abelian.

10B

311

N o w let N = k e v ( i p ) . compute that

?

T o show t h a t N is /^-invariant, l e t a e K and

(^) " = ( n
keK

e{x ) ~ ) ~ =
ak k 1 a 1

\ \ ^(x
k e K

a f c

)(

a f c

_ 1

= <p{x),

where the last equality holds because a k runs over a l l elements of K as k does. It follows t h a t < p ( x ) = t p ( x ) , and hence i f <p{x) = 1, then also < p ( x ) = 1 for a l l a e K . T h i s shows t h a t i V is AT-invariant, as claimed.
a a a

O f course, N = ker(v?) < V, so to complete the proof, we must show t h a t U n N = 1 and U N = V. If u P , then u P for a l l keK, and thus 0(u ) = (ti ) = u , and it follows t h a t cp(u) = u , where we recall that m = \ K \ . B y hypothesis, therefore, the restriction of tp to P is b o t h injective and surjective, and i n particular, U n N = C/nker(yj) = 1. F i n a l l y , we have < p ( U N ) = <p(U) = U, and thus also <p(V) = U. Since U N Z} N , it follows by the correspondence theorem t h a t U N = V, as required.
fc f c f c _ 1 fc fc_1 m

We can now prove T h e o r e m 10.15, and as we have observed, H u p p e r t ' s theorem (Theorem 10.12) w i l l then follow. P r o o f o f T h e o r e m 1 0 . 1 5 . R e c a l l t h a t P < N , where P is a nonabelian metacyclic Sylow p-subgroup of N , and p > 2. O u r goal is to show that N has an abelian h o m o m o r p h i c image of order divisible by p , and we proceed by i n d u c t i o n on \P\. Since P is metacyclic, P ' is cyclic, and of course P ' < N . A l s o P ' > 1, and thus P ' has a unique subgroup of order p , w h i c h is necessarily n o r m a l i n N . N o w let Y be an a r b i t r a r y n o r m a l subgroup of order p i n N . T h e n P / Y is a n o r m a l Sylow p-subgroup of N / Y , and it is m e t a c y c l i c by L e m m a 10.13(a). If P / Y is nonabelian, it follows by the i n d u c t i v e hypothesis that N / Y has a n abelian h o m o m o r p h i c image of order divisible by p , and thus N has such a h o m o m o r p h i c image and we are done. W e can assume, therefore, t h a t P / Y is abelian, and thus P ' C Y. It follows t h a t Y = P ' , a n d thus \P'\ - p , and P ' is the unique n o r m a l subgroup of order p i n N . Since \P'\ = p , it follows that P ' is central i n P , and thus P has nilpotence class 2. Because p > 2, T h e o r e m 4.8(a) guarantees t h a t the set V = {x P | = 1} is a subgroup of P , and of course, V < N . A l s o V is metacyclic by L e m m a 10.13(b), and since every cyclic subgroup and cyclic factor group of V has orderiat most p , it follows t h a t \V\ < p . F u r thermore, since P is not cyclic and p > 2, it follows by T h e o r e m 6.11 that P has more t h a n one subgroup of order p , and we deduce t h a t \V\ - p , and V is a n elementary abelian p-group.
2 2

Now write Z = Z ( P ) . l i V C Z , then P is i n the kernel of the conjugat i o n action of N o n V, and so N / P acts on V. T h i s is a coprime action, a n d

312

1 0 . M o r e Transfer

Theory

since P ' is an invariant subgroup of order p, it follows by Maschke's theor e m (or more precisely, by C o r o l l a r y 10.17) that we can write V = P ' x W, where W is ( W / P ) - i n v a r i a n t , and thus W < N . B u t \W\ = p , and this is a c o n t r a d i c t i o n since P ' is the unique n o r m a l subgroup of order p i n N . W e conclude that V % Z . Since P C Z , the group P / Z is abelian. A l s o , element x G P , the m a p [ - , x ] is a h o m o m o r p h i s m . have [ y , x ] = [ y , x } P = l , where the second equality T h u s x centralizes y P , and since x G P is arbitrary, all y G P , and we conclude t h a t P / Z is elementary
p 1

it follows that for each If y G P , therefore, we holds because | P ' | = p . we see t h a t y Z for abelian.
p

Now AT acts on P / Z , and P is contained i n the kernel of this action because P / Z is abelian. It follows that N / P acts on P / Z , a n d of course, this is a coprime action. A l s o V Z / Z is an invariant subgroup, and hence by Maschke's theorem, we can write P / Z = ( V Z / Z ) x ( H / Z ) , where H / Z is ( A V P ) - i n v a r i a n t , and thus P < N . A l s o V D H C V Z D H = Z , and since V Z , it follows t h a t V P , and thus \ V f \ H \ < p . W e conclude t h a t H has a unique subgroup of order p , and thus P is cyclic.
2

Now V is abelian and Z is central, and thus V Z is abelian. T h e n V Z / P , so Y Z / Z < P / Z , and thus P / Z is n o n t r i v i a l , and P is not central in P . N o w write C = C ( H ) , so t h a t P % C, and thus A T / C has order divisible by p . A l s o , N / C is isomorphically embedded i n A u t ( P ) , and since P is cyclic, A u t ( P ) is abelian, and thus N / C is an abelian homomorphic image of W w i t h order divisible by p . It follows that p divides |JV : JV'|, as required.
N

Problems

10B

1 0 B . 1 . G i v e n a prime p and an integer n > 0, let C = ( x ) be a cyclic group of order p .

n

Let a G A u t ( G ) with x

, and let P = CX ( CT ). Show

t h a t P is a metacyclic p-group w i t h nilpotence class n . H i n t . Use T h e o r e m 4.7. N o t e . T h i s problem shows that metacyclic p-groups can have a r b i t r a r i l y large nilpotence class, and we know t h a t if p > 2, such a group cannot have C p l C as a h o m o m o r p h i c image. It follows t h a t Y o s h i d a ' s theorem is s t r i c t l y stronger t h a n the assertion t h a t a Sylow normalizer controls p-transfer if the Sylow p-subgroup has nilpotence class less t h a n p.
p

1 0 B . 2 . L e t E < G, where E is an elementary abelian p-group, and suppose that | G : C ( P ) | is not divisible by p. Show that E C S o c ( G ) .
G

IOC

313

IOC If H is an a r b i t r a r y subgroup of a finite group G, then of course, the derived subgroup G' is contained i n the kernel of the transfer h o m o m o r p h i s m G - > H / H ' . I n this section we discuss another connection between G' and transfer; we consider what happens if G' plays the role of H . 1 0 . 1 8 . T h e o r e m . L e t G be a finite g r o u p . p h i s m G - > G ' / G " is the trivial map. Then t h etransfer homomor-

For applications w i t h i n group theory, we generally seek results (such as Y o s h i d a ' s theorem) that allow us to prove i n certain cases t h a t the transfer h o m o m o r p h i s m is n o n t r i v i a l . B u t T h e o r e m 10.18, does just the opposite: it shows t h a t the transfer m a p is t r i v i a l w h e n the target is the derived subgroup. There do not seem to be m a n y group-theoretic applications of this result, but it does have applications to algebraic number theory, and in p a r t i c u l a r to class field theory. (In t h a t context, T h e o r e m 10.18, w h i c h was first proved by P. Furtwangler, is usually called the " p r i n c i p a l ideal theorem".) W e have decided to present this result because it is s t r i k i n g a n d easy to state, and yet it seems quite difficult to prove. T h e proof we are about to present is essentially due to G . H o c h s c h i l d . It uses some elementary r i n g theory, and so it is quite different i n flavor from everything else i n this book, but we hope t h a t readers w i l l find it to be accessible. Let G be a finite group and let R be any ring. ( R e c a l l that by definition, rings contain u n i t y elements.) T h e g r o u p r i n g R [ G ] is the set of formal sums of the form r g , where the s u m runs over g G G , and the coefficients r lie i n R. Elements of the group r i n g are added i n the obvious way:
g g

Y
geG

9 9

+Y
g

9 9

= Y (

9 ) 9 >

eG

geG

and this makes R [ G ] into a n additive group. W e view G as a subset of R [ G ] by identifying the group element g G G w i t h the s u m r x , where r = 1 a n d r 0 for x ^ g . F r o m this point of view, an a r b i t r a r y element f > in R [ G ] is no longer just a formal sum; it is a n a c t u a l linear c o m b i n a t i o n of the elements of G w i t h coefficients i n R.
x g x 9 5

To define m u l t i p l i c a t i o n i n R [ G ) , we use the m u l t i p l i c a t i o n i n G together w i t h the m u l t i p l i c a t i o n i n R and the d i s t r i b u t i v e law. T h u s

where h = Y
x,yeG
x y = g

314

10. M o r e Transfer

Theory

It is routine to check that these definitions of a d d i t i o n and m u l t i p l i c a t i o n make R [ G ] into a ring i n w h i c h the identity element of G is the u n i t y element. We m e n t i o n that the group r i n g R [ G ] can be defined even if G is an infinite group, but i n t h a t case, the elements of R [ G ] are just those formal sums r g i n w h i c h a l l but finitely m a n y of the coefficients r are zero. Some of our results h o l d i n this more general setting, and i n fact, some of our proofs go t h r o u g h essentially unchanged. Nevertheless, we w i l l generally assume t h a t G is finite i n what follows.
g g

We shall need only the integer group ring Z[G], where Z is the o r d i n a r y r i n g of integers. In this case, where R = Z , there is a n a t u r a l connection between the group ring and actions of G on abelian groups, and we w i l l exploit this i n the proof of T h e o r e m 10.18. Suppose that M is an a r b i t r a r y a d d i t i v e l y w r i t t e n abelian group, and let G be a finite group that acts v i a automorphisms on M . If x G M and g G G , we w o u l d usually w r i t e x to denote the element of M resulting from the a c t i o n of g on x , but i n this setting, where M is w r i t t e n additively, it is more convenient to write x g i n place of x . I n this notation, we have x l = x and ( x g ) h = x ( g h ) for x G M and g , h e G . A l s o , since the given action is assumed to be v i a automorphisms, we have ( x + y ) g = x g + y g for x , y G M and g G G . If n G Z and x G M , then of course, x n is just the element that w o u l d be w r i t t e n x i n the standard m u l t i p l i c a t i v e n o t a t i o n , and since the action is v i a automorphisms, we have ( x n ) g = [ x g ) n for g G G , n G Z and x G M .
9 n

N o w let e g G Z [ G ] , where, of course, e G Z for g G G . If x G M , we can define a "right action" of the r i n g Z[G] on M by setting
g g

It is routine to check for a , 3 G Z[G] and x , y G M that (a) x { a + (3) = x a + x 0 , (b) ( x a ) p = x(a(3),

(c) (x + y ) a = x a + y a and, of course (d) x l = x , where 1 is the u n i t y of Z [ G ] . In other words, M is a right Z[G]-module. O f course, if we had started w i t h a "left action" v i a automorphisms of G on M , so that g ( h x ) = [ g h ) x for g , h G G a n d x G M , the analogous construction w o u l d make M into a left Z [ G ] , module. T e m p o r a r i l y p u t t i n g modules aside, we define the a u g m e n t a t i o n m a p 6 : Z[G] -> Z by setting

IOC

315

so t h a t 8 ( a ) is s i m p l y the s u m of the coefficients appearing i n a G Z [ G ] . It is routine to check that 8 is a ring h o m o m o r p h i s m , a n d so its kernel, denoted A ( G ) , is a n ideal of Z[G]; it is called the a u g m e n t a t i o n ideal. 10.19. L e m m a . L e t G be a finite g r o u p . T h e n A ( G ) i s t h e a d d i t i v e s u b g r o u p of Z [ G ] g e n e r a t e d by t h e elements g - 1 f o r 1 / g G G, a n d i n f a c t , these g e n e r a t o r s f o r m a Z-basis for A(G). W e should e x p l a i n what we m e a n by a "Z-basis". L e t M be an a d d i tively w r i t t e n finitely generated abelian group, and let { x x , . . . , x } be a generating set. T h e n the subset of M consisting of elements of the form J2 e i X r for e G is clearly a subgroup of M that contains a l l of the generators x . It is thus a l l of M , and hence every element of M is a linear c o m b i n a t i o n of the f o r m ^ e ^ . T h e given generating set is a Z -basis for M if the only way to get = 0 is for a l l coefficients e = 0. I n this
u 2 n %

case, each element of M is u n i q u e l y of the form e x . F o r example, if G is a finite group, then G is a Z-basis for the additive group of Z \ G \ .
% %

Of course, not every finitely generated a b e l i a n group has a Z-basis; those that do are said to be free abelian groups. A n o n t r i v i a l finite abelian group, for example, cannot have a Z-basis, and so it is definitely not free abelian. In fact, one can define infinite Z-bases analogously, and the conclusion of L e m m a 10.19 holds for infinite groups too. P r o o f of L e m m a 10.19. Certainly, 8 ( g - 1) = 0 for g G G , where 8 is the augmentation map, and thus g - 1 G k e r ( i ) = A ( G ) . N o w let a be an a r b i t r a r y element of A ( G ) , and w r i t e a = e g, w i t h e G Z. T h e n
g g

and since e = 8 ( a ) = 0, we see t h a t a is a Z-linear c o m b i n a t i o n of the elements g - 1 for 1 / g G G , and hence these elements form a generating set for A ( G ) .
g

To check that these generators form a coefficients f

g

Z-basis,
g

suppose t h a t there exist

G Z for 1 + g G G such t h a t

f ( g - 1) = 0. T h e n

and since G is a

Z-basis

for

Z[G],

it follows t h a t the coefficient f

of g is

zero for a l l nonidentity elements g e G .

u 2 n

Suppose t h a t { x x , , x } is a Z-basis for some a d d i t i v e l y w r i t t e n abelian group M , and let A be an a r b i t r a r y ( m u l t i p l i c a t i v e ) abelian group.

316

10. M o r e Transfer

Theory

Choose a r b i t r a r y elements a* A for 1 < i < n , a n d define the m a p 9 : M - A by setting 9(J2e e

l X i

J]K)ei-

N o t e t h a t 9 is well denned since every element of M is u n i q u e l y of the form

i X l

w i t h e G Z, a n d it is t r i v i a l to check t h a t 9 is a group h o m o m o r p h i s m .
{

T h i s shows that i f M is a free a b e l i a n group a n d A is any a b e l i a n group, t h e n there is h o m o m o r p h i s m 9 : M - A such t h a t 9 carries the members of a

Z-basis

for M to a r b i t r a r i l y selected elements of A .

N o w let M be a left ^ - m o d u l e , a n d l e t X C R a n d Y C M be additive subgroups. ( T h e following remarks also a p p l y i n the i m p o r t a n t case where M = R, a n d where X a n d Y are additive subgroups of R . ) R e c a l l that is the set of a l l (finite) sums a n d differences of elements (Equivalently, XY is the by definition, XY

of the form xy i n M , where x G X a n d y G Y.

additive subgroup of M generated by a l l products of the form x y . ) Observe t h a t i f X a n d Y are finitely generated (as additive groups) by subsets { x }
r

and { y } respectively, t h e n XY
3

is finitely generated by the products

2

xy.
l j

In p a r t i c u l a r , i n Z[G], the a u g m e n t a t i o n ideal A ( G ) is an additive subgroup generated by the elements g - 1 for g G G , a n d thus A ( G ) is the additive subgroup generated by elements of the form (g - l ) ( h g , h G G . N o t e that A(G)
2

1) for

A(G)Z[G]

C A ( G ) since A ( G ) is a n ideal, a n d thus

C A(G).

1 0 . 2 0 . T h e o r e m . L e t G be a additive group A(G)/A(G) the coset ( g - 1) + A ( G ) .

2 2

finite

g r o u p . Then

G/G'

i s i s o m o r p h i c t o the G'g t o

v i a a n i s o m o r p h i s m c a r r y i n g the coset

P r o o f . Define <p : G

A(G)/A(G)
2

by setting <p(g) = (g - 1) + A ( G ) . + (x - l ) ( y - 1), a n d a n d hence

2

For elements x, y G G , we have xy-

1 = ( x - 1) + (y-1)

since ( x - l ) ( y - 1) G A ( G ) , it follows that c p ( x y ) = <p(x) + <p(y),

ip is a h o m o m o r p h i s m . A l s o , since the additive group A ( G ) is generated by elements of the form x - 1 w i t h x G G , we see t h a t A ( G ) / A ( G )

2

is

generated by the cosets (x - 1) + A ( G ) , a n d thus <p(G) is a generating set for A ( G ) / A ( G ) .

2

B u t <p(G) is a subgroup, a n d thus it must be the

2

whole group A ( G ) / A ( G ) , a n d hence <p is surjective.

A l s o , since <p is a

h o m o m o r p h i s m from G to an a b e l i a n group, it follows t h a t G' C ker(y>). T o complete the proof, we show t h a t i n fact, G' = ker(</?), a n d thus <p defines an i s o m o r p h i s m from G/G' onto A ( G ) / A ( G ) .
2

N e x t , we construct a m a p i n the opposite d i r e c t i o n . B y L e m m a 10.19, the set { g - l \ 1 ^ g G G } is a

Z-basis for

A ( G ) , a n d since G/G' G/G'

is a n a b e l i a n

group, there is a h o m o m o r p h i s m 9 : A ( G )

such t h a t 9 ( g - 1) = G'g

for 1 ^ g G G . ( O f course, this formula also holds w h e n g = 1 since 9 is a h o m o m o r p h i s m , a n d thus 9 ( 0 ) is the identity element of G / G ' . )

IOC

317

Since 0 is a h o m o m o r p h i s m from the additive group A ( G ) to G / G ' , and % - 1) = G ' g for a l l elements g E G, we can a p p l y 0 to b o t h sides of the identity xy - 1 = ( x - 1) + (y - 1) + (x - l ) ( y - 1) to get G ' ( x y ) = (G'x)(G'y)6((x-l)(y-l)) for elements x , y G G . B u t G ' ( x y ) = ( G ' x ) ( G ' y ) , and it follows that 0 ((x - l ) ( y - 1)) is the identity, and thus (x - l ) ( y - 1) G ker (0). T h e elements (x - l ) ( y - 1) generate A ( G ) as a n abelian group, however, and thus A ( G ) C ker (0).
2 2

We have seen that ker(v?) D G' a n d t h a t it suffices to prove equality here. T o establish the reverse containment, let k E k e r ( p ) . Since <p(k) is the 0 element of A ( G ) / A ( G ) , and by definition, i p ( k ) = ( k - 1) + A ( G ) , it follows t h a t k - l E A ( G ) C ker (0), and thus G'k = 6 ( k - 1) is the identity of G / G ' , and hence k G G ' , as required.
2 2 2

Let K C G be a subgroup, and view Z [ K ] C Z[G]. L e t T be a right transversal for K i n G , so that each element g G G is uniquely of the form k t , where k f] K and t ET. If a G Z[G] is arbitrary, therefore, we can write

where e G and a = f c c G Z [ K ] . W e refer to a as the i - c o m p o n e n t fi of a w i t h respect to T, and we observe t h a t for each element t G T, the m a p f : a ^ a from Z[G] to Z [ K ] is a well defined h o m o m o r p h i s m of additive groups. W e shall need the following technical result.
fct t t t t

eyfc

1 0 . 2 1 . L e m m a . L e t K C G, w h e r e G i s a finite g r o u p , a n d l e t T be a right t r a n s v e r s a l f o r K i n G such t h a t 1 G T. W o r k i n g i n Z [ G ] , l e t a G A ( K ) A ( G ) , a n d f o r t G T, w r i t e a t o d e n o t e t h e t - c o m p o n e n t of a w i t h r e s p e c t t o T. T h e n a G A ( K ) a n d J 2 t E A(K) .

t a 2 t t

P r o o f . L e t f be the map c a r r y i n g a n element of Z[G] to its i-component w i t h respect to T, and w r i t e / = Y \ f . T h e n / and a l l of the maps f are additive homomorphisms, and our goal is to show that f ( A ( K ) A ( G ) ) C A ( K ) for a l l t E T, and that / ( A ( K ) A ( G ) ) C A ( K ) . Since the elements ( k - l ) ( g - 1) for k E K and g E G generate the additive group A ( K ) A ( G ) , it suffices to prove the l e m m a i n the case a = ( k - l ) ( g - 1).
t t t t 2

F i r s t , suppose t h a t g E K . T h e n a = ( k l ) ( g 1) G Z [ K ] , a n d thus the 1 -component of a is a , a n d a l l other components are zero. Since a E A(K) C A ( K ) i n this case, there is n o t h i n g further to prove.
2

We can now assume that g K , and we w r i t e g = hs, where h E K and s E T {1}. T h e n a = (kl ) ( g - 1) = ( k - l ) g - ( k - 1) = ( k - l ) h s - ( k - 1 ) 1 ,

318

1 0 . M o r e Transfer

Theory

and thus a = ( k l ) h G A ( K ) a n d a = - ( k 1) G A ( K ) . A l l other components of a are zero, a n d thus a l l components lie i n A(AT), as required. 2 A l s o , f ( a ) = ( k l ) h - (fc - 1) = (fc - l ) ( h 1) G A(AT) , a n d the proof is complete.
s x

10.22. C o r o l l a r y . L e t K C G, w h e r e G i s a finite g r o u p . T h e n i n Z[G], w e h a v e A ( K ) = A(AT)A(G) n A { K ) = A(AT)A(G) n Z [ K ] .

2

P r o o f . It is clear that A(AT) C A(AT)A(G) n A(AT) C A ( K ) A ( G ) n Z [ K ] , 2 so it is sufficient to show that A ( K ) A ( G ) n Z [ K ] C A(AT) . L e t T be as i n L e m m a 10.21, a n d observe t h a t i f a G A ( K ) A ( G ) n Z [ K ] then the 1component of a is a itself, a n d a l l other components are zero. T h u s a is the 2 sum of its components, a n d so a G A(AT) b y L e m m a 10.21.
2

If K C

A(G).

O f course

G, then since A(G) is a n ideal A(AT)A(G) is a n o r m a l

of

Z[G],

we have

A(AT)A(G)

subgroup of the additive group

A(G), a n d i n the next several results, we study the factor group A(G) =

A(G)/A(AT)A(G).

W e use the standard "bar convention", so that i f A is A(G). I n p a r t i c u l a r , since A { K ) C A(G), we see

any additive subgroup of A(G), then A is the image of A under the canonical h o m o m o r p h i s m A(G) t h a t A (AT) is a subgroup of A(G). 10.23. C o r o l l a r y . L e t K C G, w h e r e G i s a finite g r o u p . W o r k i n g i n Z [ G ] , let

A(G) A(G) A(AT)A(G) '

T h e n A(AT) = AT/AT', w h e r e k 1 maps t o K ' k f o r elements k G AT. P r o o f . T h e m a p k - l ^ k - l defines a h o m o m o r p h i s m from A (AT) onto A(K) k~^~l know kernel is equal to

A(AT) n A(G)A(AT). B y C o r o l l a r y 10.22, this A(AT) , a n d thus A ( K ) ^ A(AT)/A(A/) v i a the m a p H-> (fc - 1) + A (AT) . B y T h e o r e m 10.20 a p p l i e d to the group AT, we that A(AT)/A(AT) ^ AT/AT' v i a the m a p (fc - 1) + A(AT) ^ AT'fc.
w i t h kernel equal to
2 2 2 2 2

T h e m a p i n the statement of the corollary is the c o m p o s i t i o n of these two isomorphisms.

W e are interested i n the transfer h o m o m o r p h i s m v : G - > K / K ' , where K V

G.
:G

Since K / K '

g A (AT)

i n the n o t a t i o n of C o r o l l a r y

10.23, we c a n

ask

w h i c h subgroup of A(AT) corresponds to v ( G ) under this i s o m o r p h i s m . If K is a pretransfer m a p , then v ( G ) is the image of V ( G ) i n K / K ' . T h e subgroup of A (A/) corresponding to v ( G ) , therefore, is exactly the set of elements of the form V ( g ) - 1 for g G G. W e shall be able to give a useful description of this subgroup i n the case where K is n o r m a l i n G.

IOC

319

A s s u m e now that AT < G. If g e G, then conjugation by g defines an a u t o m o r p h i s m of AT, and hence g - A ( K ) g = A (AT). T h e n A ( K ) g = g A { K ) , a n d since g A ( G ) = A ( G ) , we have g A ( K ) A ( G ) = A ( K ) g A ( G ) = A ( A / ) A ( G ) , and thus the additive group A ( A T ) A ( G ) is invariant under left m u l t i p l i c a t i o n by elements of G . T h i s subgroup of Z [ G ] , therefore, is a left ideal. (It is also a right ideal, but we shall not need t h a t fact.) O f course, A ( G ) is also a left ideal, and thus A ( G ) = A ( G ) / A ( A / ) A ( G ) is a left Z [ G ] module. Clearly, A ( G ) is a n n i h i l a t e d by left m u l t i p l i c a t i o n by elements of A(AT), and thus ( k - l ) a = 0 for a e A ( G ) and k K . T h e n k a = a , and since the a c t i o n of K is t r i v i a l , we have a left a c t i o n of G / K on A ( G ) .
l

N o w let T be a transversal for K i n G , a n d note that since we are assuming t h a t AT is n o r m a l , there is no d i s t i n c t i o n between left and right transversals. L e t a e Z[G] be the s u m of the elements of T, a n d observe t h a t left m u l t i p l i c a t i o n by a defines a h o m o m o r p h i s m of additive groups S : A ( G ) -> A ( G ) . A l s o , since left m u l t i p l i c a t i o n by elements of AT fixes a l l elements of A ( G ) , the m a p 3 is independent of the transversal T. 1 0 . 2 4 . T h e o r e m . L e t v . G ^ K / K ' be t h e t r a n s f e r h o m o m o r p h i s m , G i s a finite g r o u p a n d K < G. W o r k i n g i n Z [ G ] , w r i t e X7cT_
1 ;

where

A(G) A(A~)A(G)' by left m u l t i p l i c a t i o n by t h e

a n d l e t E : A ( G ) -> A ( G ) be t h e m a p d e f i n e d sum of t h e elements

of a t r a n s v e r s a l f o r K i n G. T h e n v { G ) 9* S ( A ( G ) ) .

P r o o f . L e t V : G -> AT be a pretransfer m a p , so t h a t

V(g) = l [ t g ( t - g ) ter

where T is some transversal for AT i n G , and the p r o d u c t is taken i n some fixed but unspecified order. T h e n V ( g ) e K , and s o V ( j ) - l A (AT). U n d e r the i s o m o r p h i s m between A(AT) a n d K / K ' i n C o r o l l a r y 10.23, this element corresponds to the image of V ( g ) i n K / K ' , w h i c h is v { g ) . W e argue next that V ( g ) - l To see this, w r i t e t g = k ( t - g )
t

= E ( g - l ) .
t

for each element t e T, where k

AT. is a

Then

= ]"]>*, and since by C o r o l l a r y 10.23, the m a p fc ^

fc^T

h o m o m o r p h i s m from K to the additive group A (AT), we have V { g ) - l = Y ter

k d

t ~

320

1 0 . M o r e Transfer

Theory

A l s o , recall t h a t 3 ( a ) = era = era, where a = Y \ t, and thus 3(^=1) = We have i f o - l ) .

g*c^

-1) = 53*^ - 53*

k {t-g)
t

~ Y , t - 9
t e r

teT

= J2(k
teT

l){t-g)

= Y , ( k - l )
t

mod A ( K ) A ( G ) ,

where the final congruence holds because ( k - l ) x = ( k - l ) m o d A ( A T ) A ( G ) for a l l x G and A; AT. T h u s 3(^1) = 5 3 ^ 1 as c l a i m e d . Now A A(K) let X be the image of the m a p g ^ V ( g ) - 1 from G to A(AT). T h e n T h e map 3 is a corresponds to v ( G ) under the i s o m o r p h i s m of L e m m a 10.23 between and AT/AT', and it follows that A is a subgroup. = ^ ) - ! .

h o m o m o r p h i s m from A ( G ) into itself, and we have shown that 3 carries the generators of A ( G ) onto the subgroup X . It follows t h a t 3 ( A ( G ) ) = A = v ( G ) , as wanted.

R e c a l l t h a t the assertion of T h e o r e m 10.18 is t h a t the transfer homom o r p h i s m v : G -> G ' / G " is always t r i v i a l . A c t u a l l y , something a little more general is true. 1 0 . 2 5 . T h e o r e m . L e t G be a f i n i t e g r o u p , a n d l e t v : G transfer h o m o m o r p h i s m , elements g e G . w h e r e G' C AT C G . T/ien w ( # ) l
K:G

AT/A"

be t h e

' l = 1 / o r aiZ

Of course, if AT = G ' , then T h e o r e m 10.25 asserts t h a t v ( g ) = 1 for all g G , and so T h e o r e m 10.18 is an immediate consequence. T o prove T h e o r e m 10.25, we need the following general result about modules for comm u t a t i v e rings. 1 0 . 2 6 . T h e o r e m . L e t A be a left R - m o d u l e , w h e r e R i s a c o m m u t a t i v e r i n g , a n d l e t U C R be a n i d e a l . Assume t h a t A a n d U a r e finitely g e n e r a t e d as

IOC

321

a d d i t i v e g r o u p s , a n d suppose exists a n element r G R such

t h a t U A has

finite

index m i nA .

Then there

t h a t r A = O a n d r = m - l m o d U.

We review a bit of m a t r i x theory before we begin the proof. If S is an n x n m a t r i x over the c o m m u t a t i v e r i n g R, the c l a s s i c a l a d j o i n t T of S is the n x n m a t r i x whose ( i , j ) - e n t r y is plus-or-minus the determinant of the ( j , i ) - m i n o r of S, where the sign is . T h e fact about the classical adjoint t h a t we need is t h a t TS = d l , where d = det(S) identity m a t r i x . and I is the n x n

P r o o f o f T h e o r e m 1 0 . 2 6 . B y the fundamental theorem of abelian groups, the finite abelian group A / U A of order m is a direct p r o d u c t of cyclic subgroups d for 1 < i < k. W r i t e m , = \ C \ , and observe t h a t m = \ A / U A \ = W r r n . N o w choose elements a G A such t h a t a generates d m o d u l o U A , a n d let B = fa | 1 < * < k ) , so t h a t A = B + U A .
X { {

N e x t , we e x p a n d the set { a , | 1 < i < k } to o b t a i n an appropriate generating set for A . T o do this, we observe t h a t U A is finitely generated as a n additive group since b o t h U and A are finitely generated, a n d so we can choose a finite generating set X for U A . Choose n o t a t i o n so that X = { | k < i < n } , where n = k + \ X \ , a n d observe t h a t since A = B + U A , the set { | 1 < i < n } generates A . L e t = 1 for k < i < n and note t h a t n ? = i m = m and r m a i G U A for 1 < i < n .
a i a i

Since mm G U A for 1 < i < n and A is generated b y the elements for 1 < j < n , we can write

3=1 where s G /, and we let S be the n x n m a t r i x over R w i t h ( i , j ) - e n t r y equal to ' . A l s o , let M be the diagonal n x n m a t r i x w i t h (t, i)-entry equal t o m l , and finally, let v be the n x 1 c o l u m n vector w i t h entry i equal to en. W e thus have M v = Sv, and thus ( M - S)v = 0.
i d S i J r

Now let T be the classical adjoint of the n x n m a t r i x M - 5 over i?. T h e n T ( M - S) = r l , where r = d e t ( M - 5 ) and / is the n x n identity m a t r i x over R, and we have 0 = T ( M - S)v = r l v = r v a n d thus r a = 0 for 1 < i < n . Since the elements a generate A , we have r A = 0, as required. A l s o , t h e entries of S lie i n U, so we have
{ {

r = d e t ( M - S) = d e t ( M ) = and the proof is complete.

f[ m i

! = m - l

mod U

322

10. M o r e Transfer

Theory

We A(G)

r e t u r n now to group rings of finite groups. R e c a l l that i i K < G = A { G ) / A ( K ) A ( G ) is a left Z[G]-module. finite g r o u p . I n 1\G\, l e t

then

10.27. L e m m a . L e t K < G, w h e r e G i s a

A(K)A(G) ' m u l t i p l i c a t i o n by t h e Suppose t h a t e e Z[G] map.

a n d l e t E : A ( G ) -> A ( G ) be t h e map i n d u c e d by left sum Then of t h e elements of a t r a n s v e r s a l f o r K i n G. satisfies

e A ( G ) = 0, a n d l e t m = 5 { e ) , w h e r e 5 i s t h e a u g m e n t a t i o n

\ G : K \ d i v i d e s m, a n d ( m / \ G : K \ ) > E ( A ( G ) ) = 0.

P r o o f . L e t T be a transversal for K i n G w i t h 1 6 T , a n d let a = Y \ t, so t h a t S ( a ) = < r a = ra for a g A ( G ) . G i v e n t 6 T, we k n o w that the left-multiplication maps o n A ( G ) by a l l elements of the coset K t are equal, and thus it is no loss to assume that
e

= J2
ter

e t t

for some integers e w i t h V \ e = m. have

t t

If g G , then since e A ( G ) = 0, we mod A ( K ) A { G ) .

(J2ett)(g-l)
teT

= 0

Now

for t 6 T, write t g = k { t - g ) , where k G A". T h e n

t t

(E
teT

a t ) ( g - l ) = J2 e t h ( t - g ) - J2
teT t e T

e t t

teT
t t t

teT

= ^{e k -e .gl){t>g),
teT

and by the previous paragraph, this element lies i n A ( A - ) A ( G ) . N o w the (i-o)-component of this element w i t h respect to T is e k - et.pl, a n d by L e m m a 10.21, this component must lie i n A ( K ) . It follows t h a t e - e . = 0 for a l l t T and g G.
t t t t g

Since the dot action of G is transitive o n T , the numbers e must a l l be equal, say to e, and thus e = ea. A l s o , m = V > i = e\T\ = e\G : K \ , a n d thus | G : K \ divides m , as wanted. F i n a l l y , i f a 6 A ( G ) , we have
t

(m/\G:K\)-E(a) and the proof is complete.

= eaa = ea = 0,

IOC

323

P r o o f o f T h e o r e m 1 0 . 2 5 . A s usual, write A ( G ) = A{G)/A(K)A(G), and observe t h a t AT < G since G' C AT. L e t E : A ( G ) > A ( G ) be the (uniquely determined) map induced by left m u l t i p l i c a t i o n by the s u m i n Z[G] of the elements of an a r b i t r a r y transversal for AT i n G. B y Theor e m 10.24, the transfer image v ( G ) is isomorphic to E ( A ( G ) ) , and so to show that v { g ) \ - ' \ = 1 for all g i n G, it suffices to show t h a t the m a p | A : G ' \ E is identically zero on A ( G ) .
K G

W r i t e A - A ( G ) , and recall that AT acts t r i v i a l l y (by left m u l t i p l i c a t i o n ) on A , and thus the abelian group G / K acts on A by left m u l t i p l i c a t i o n , and so A is n a t u r a l l y a left module for the c o m m u t a t i v e r i n g R = Z [ G / K ] . W e want to a p p l y T h e o r e m 10.26 w i t h U = A ( G / K ) , and so we must compute the index \ A : U A \ . W e argue t h a t \ A : U A \ = \ G : G'\. T o see this, observe that the a c t i o n of an element K g e G / K on A is just the action of g on A , and thus UA = A ( G / K ) A = E
KgeG/K

((Kg)-1)A

= J 2 ( 9 - 1 ) A = &(G)A.
geG

Also, A ( G ) A = A(G)A(G) = and thus \ A : U A \ = \ A j G j : A ( G p | = | A ( G ) : A ( G ) | = \ G : G'\, where the second equality holds because A ( A ) A ( G ) C A ( G ) , and the t h i r d holds by T h e o r e m 10.20. T h e o r e m 10.26 applies since A and A ( G / K ) are finitely generated, and it follows t h a t there exists an element 7 6 Z[G/AT] such t h a t A = 0 and 7 = \ G : G ' l - l m o d A ( G / K ) . Observe t h a t this congruence m o d u l o A ( G / K ) implies that the augmentation of 7 i n the group r i n g Z [ G / K ] is equal to |G:G'|.
7 2 2

A{GY,

Since the element K g of G / K acts on A as g does, we can use 7 to construct an element e G Z[G] that acts on A as 7 does, and for w h i c h the augmentation is equal to the augmentation of 7, namely \ G : G'\. T h u s eA = 0 and 5(e) = \ G : G ' | , and therefore L e m m a 10.27 guarantees t h a t ( | G : G ' | / | G : AT|)S is identically zero on A ( G ) . Since | G : G ' | / | G : K \ = \ K : G'\, we are done. W e close w i t h an a p p l i c a t i o n of T h e o r e m 10.18, o r i g i n a l l y proved by J . A l p e r i n and T . K u o w i t h a different proof. 1 0 . 2 8 . C o r o l l a r y . L e t G be a g\G:A\

_ 1

j nl f
Qr a

finite

group,

a n d l e t A = G' n Z ( G ) .

Then

e emen s

g ^ G .

324

10. M o r e Transfer

Theory

P r o o f . L e t V : G - > A , U : G -> G' and W : G' -> A be pretransfer maps, and let g e G. Since A is abelian, V is a c t u a l l y the transfer h o m o m o r p h i s m from G to A and is the transfer h o m o m o r p h i s m from G' to A . N o w A C Z ( G ) , and thus = by T h e o r e m 5.6. B y T h e o r e m 10.8, w h i c h is the t r a n s i t i v i t y of transfer theorem, and again using the fact t h a t A is abelian, we have V { g ) = W { U { g ) ) . N o w T h e o r e m 10.18 tells us t h a t U { g ) G G", and thus U ( g ) G k e r ( W ) . W e now have g \ \ = V ( g ) = W ( U ( g ) ) = 1, as wanted.
a A

Problems IOC
1 0 C . 1 . L e t <p : G -> H be a group h o m o m o r p h i s m . be extended to a r i n g h o m o m o r p h i s m 9 : Z [ G ] A(N)Z[G] = k e r ( 9 ) = Z [ G } A { N ) , where N = k e v { < p ) . Show t h a t ip can Z [ H ] , and show t h a t

1 0 C . 2 . L e t A be a n a d d i t i v e l y w r i t t e n finitely generated free abelian group. Show t h a t a l l Z-bases for A have the same finite size. H i n t . L e t p be a prime number and let B = p A . Show t h a t A / B dimensional vector space over the field F of order p . is a finite

1 0 C . 3 . L e t i f be a subgroup of the group U of invertible elements i n Z [ G ] , and suppose t h a t I f is a Z-basis for the additive group of Z[G]. Show that there is a subgroup K C U such that K ^ H a n d K is also a Z-basis for Z [ G ] , and such t h a t 8 { k ) = 1 for a l l elements k K , where 8 is the augmentation map on Z [ G ] . N o t e . It had been conjectured that i n this situation, H and G were necessarily isomorphic, but this is now k n o w n to be false, 1 0 C . 4 . A s i n the previous problem, let H be a Z-basis for Z [ G ] , where H is a subgroup of the group U of invertible elements i n Z [ G } . Show that H/H'. GIG' ^

1 0 C . 5 . L e t | G | = n and suppose t h a t the elements of G are numbered, so t h a t G = {01, g , . . . , g } - G i v e n an element a Z [ G ] , let M { a ) be the n x n m a t r i x over Z whose (i, j ) - e n t r y is equal to the coefficient of g i n a . Show t h a t the trace t r ( M ( a ) ) is ne, where e G Z is the coefficient of 1 i n a .
2 n 5 9 i

1 0 C . 6 . In the situation of the previous p r o b l e m , suppose t h a t a = 1 for some integer m . Show t h a t a l l eigenvalues of the m a t r i x M { a ) are roots of unity, and deduce t h a t e G { - 1 , 0 , 1 } . H i n t . Show t h a t M ( a ) is the identity m a t r i x , and use the fact t h a t the trace of an n x n m a t r i x is equal to the s u m of its n eigenvalues, counting multiplicities.
m

Appendix: The Basics

In this appendix, we provide a quick review of some of the basic facts of elementary group theory t h a t are assumed i n the foregoing chapters. T h e presentation here is neither detailed nor exhaustive, however, and readers may w i s h to consult the appropriate sections of a general abstract-algebra text for a more comprehensive treatment of some of this m a t e r i a l . (Parts of our b o o k A l g e b r a : A G r a d u a t e C o u r s e , for example, w o u l d be a good supplementary source for the theory needed here.) A g r o u p is a set G together w i t h a n associative b i n a r y operation on G t h a t satisfies two further conditions: there must be a (two-sided) i d e n t i t y element i n G, and each element of G must have a (two-sided) inverse w i t h respect to t h a t identity. W e generally w r i t e our groups multiplicatively, denoting the i d e n t i t y by the s y m b o l " 1 " , a n d w r i t i n g " a T " for the inverse of an element x 6 G. (It makes sense to fix this n o t a t i o n because, as is easy to check, a group G has a unique identity, and each element x e G has a unique inverse.)
1

A group t h a t plays a fundamental role i n the theory is the s y m m e t r i c g r o u p on a nonempty set X . T h i s group, denoted S y m ( A ) , is the set of all bijections from X onto itself, and the b i n a r y " m u l t i p l i c a t i o n " i n S y m ( X ) is function c o m p o s i t i o n . (The elements of S y m ( A ) are also referred to as p e r m u t a t i o n s of A . ) T h e group identity i n S y m ( A ) is the identity m a p on A , and this explains w h y the i d e n t i t y of a group is called the "identity" element. (Note t h a t i n this book, we compose functions from left to right, so that i f a , P S y m ( A ) , then a ( 3 means "do a , t h e n /?", and so we can write ( x ) ( a / 3 ) = { { x ) a ) ( 3 . ) In the case where A = { 1 , 2 , . . . , n } we write S for S y m ( A ) , and we note t h a t \ S \ = n \ .
n n

325

326

Appendix:

The

Basics

A t r a n s p o s i t i o n i n S y m ( X ) is a p e r m u t a t i o n t h a t interchanges two points, leaving a l l other members of X fixed, and it is not hard to see t h a t every p e r m u t a t i o n is a p r o d u c t of transpositions. Less t r i v i a l is the fact t h a t i f a G S y m ( X ) is w r i t t e n as a p r o d u c t of r transpositions, then a l t h o u g h the integer r is not uniquely determined by a , its p a r i t y is. In other words, either every representation of a as a p r o d u c t of transpositions involves a n even number of transpositions, or else every such representation involves an o d d number of transpositions. E x a c t l y half of the permutations on a set X containing at least two points are products of a n even number of transpositions, and these even permutations form a group called the a l t e r n a t i n g g r o u p on X , denoted A l t ( X ) . If X = {1, 2 , . . . , n } , we write A for A l t ( X ) , and we note t h a t \ A \ = (n!)/2.
n n

Now let x G G, where G is an a r b i t r a r y group. If n is an integer, we define x as follows. If n > 0, then x = xx x is the p r o d u c t of n copies of x ; i f n < 0, then, of course, - n is positive, and we define x = ( x ) " " , and finally, we set x = 1. W i t h these definitions, it is easy to check that x x = yjn-\-n gj^oj ^ x ) x for all integers m, n G Z .
n n n - 1 m n m n m n

In group theory, the word "order" has two different, but not entirely unrelated meanings. T h e order of a group G is its cardinality, denoted \ G \ , and the order of an element x G G, denoted o ( x ) , is the smallest positive integer n such that x = 1. (If there is no such positive integer, we say t h a t x has infinite order.) If o { x ) = n , then for integers a and b, we have x = x i f and only if a = b m o d n , and thus there are precisely n different elements of G t h a t are powers of x , and it follows that \ G \ > o ( x ) . If x has infinite order, then the only way to get x = x is for a = b, and so i n this case, a l l of the elements x w i t h a G Z are distinct. In particular, if x has infinite order, there are infinitely m a n y different powers of x , and thus G must have infinite order.
n a b a h a

A s u b g r o u p of a group G is a subset H C G such t h a t H is a group i n its o w n right, using the m u l t i p l i c a t i o n i n G. If H is a subgroup, it is easy to check t h a t its identity must be the i d e n t i t y of G, and the inverse i n H of a n element x G H is the inverse of x i n G . A subgroup of G, therefore, is a n o n e m p t y subset t h a t is closed under m u l t i p l i c a t i o n and inverses, and it is easy to see that conversely, every such subset is a subgroup. If x G G has finite order n > 1, then x ~ = x ~ is a power of x , and so if G is finite, it follows that it suffices to check t h a t a nonempty subset is closed under m u l t i p l i c a t i o n i n order to establish t h a t it is a subgroup; closure under inverses then follows automatically.
l n x

The p r i m a r y objects of interest i n group theory are the subgroups of a given group G, so it is almost always the case t h a t w h e n we mention some

Appendix:

T h e Basics

327

subset of G , we are t h i n k i n g of a subgroup. If we w r i t e "let H C G " , therefore, then although we may neglect to say so, we generally intend t h a t H should be a subgroup, and we do not use any special n o t a t i o n to denote subgroups. (For clarity, however, especially i n statements of theorems, we w i l l often not rely on this default, and we w i l l say e x p l i c i t y t h a t H is subgroup.) In situations where we want to consider subsets t h a t may not be subgroups, we indicate t h a t by w r i t i n g something like "let X C G be a subset". O f course, a set that we c o n s t r u c t should not a u t o m a t i c a l l y be assumed to be a subgroup. F o r example, given subsets X , Y C G , the subset X Y is defined by X Y = { x y \ x G X , yY}. B u t even i f X and Y are subgroups, it is not generally true that X Y is a subgroup. X . l . L e m m a . L e t H a n d K be s u b g r o u p s s u b g r o u p if a n d o n l y if H K = K H . of a g r o u p G. Then H K is a

P r o o f . If H K = K H , then ( H K ) ( H K ) = H ( K H ) K = H ( H K ) K = (HH)(KK) = H K , and thus H K is closed under m u l t i p l i c a t i o n . A l s o , if h G H and k G K , then ( h k ) ' = k ' ^ h ' G K H = H K , and hence H K is closed under inverses. Since 1 G H K , we see t h a t H K is nonempty, and thus it is a subgroup. Conversely, suppose t h a t H K is a subgroup. T h e n K C H K and H C H K , and since H K is closed under m u l t i p l i c a t i o n , we have K H C H K . T o prove the reverse containment, let x G H K . Since H K is closed under inverses, we have x ~ G H K , and thus we can w r i t e ~ = hk, w i t h h G H and k G AT. T h e n x = (/ifc)" = A T / * G K H , and this completes the proof.
1 1 l l 1 1 - 1 x

T h e r e is a simple formula for the c a r d i n a l i t y of H K . (We do not say "order" since H K m a y not be a group.) X . 2 . L e m m a . L e t H a n d K be finite s u b g r o u p s of a g r o u p G. T h e n \ H K \ = \H\K\/\H n K\. P r o o f . W r i t e D = H n K . T h e set H x K of ordered pairs ( h , k ) w i t h he H and k e K has c a r d i n a l i t y | / f ||AT|, and we have a m a p 6 : H x K ^ H K defined by 0((/i,fc)) = /ifc. T o complete the proof, it suffices to show t h a t each element x G H K has exactly | D | preimages w i t h respect to 9. T o see this, w r i t e x = h k w i t h h G H and jfc G K . If d D , then also ( T G D since b o t h 7f and K are closed under inverses, and thus ( h d , d ~ k ) <E H x K and 9 { { h d , d - k ) ) = h k = x . T h e pairs { h d , d ~ k ) are distinct for distinct elements d G D , and this yields | D | preimages for x . T o see that we have all possible preimages, suppose t h a t 9 ( ( h ' , jfc')) = x w i t h h ' H and jfc' G AT. T h e n h k = x = ft'jfc', and thus jfc(jfc')" = h ^ t i , and this element, w h i c h we call d, lies i n H n AT = >. W e have / i f t ' = d, so h ' = hd, and also,
1 l l l 1 _ 1

328

Appendix:

The

Basics

d = k ( k ' ) - \ so k' = d~ k. In other words, ( h ' , k ' ) is one of the preimages t h a t we have already constructed, and the proof is complete. The following is another useful fact about products of two subgroups.

X . 3 . L e m m a ( D e d e k i n d ) . L e t H a n d K be s u b g r o u p s of a g r o u p G, a n d l e t H C U C G, w h e r e U i s a l s o a s u b g r o u p . T h e n H K n U = H ( K n U ) . P r o o f . It is clear t h a t H ( K n t / ) C H K , and since H C U, we also have H { K n U ) C U. T h u s H ( K n U ) C FAT n U, and it suffices to prove the reverse containment. L e t u G H K n t/, and write u = hk, w i t h ft e 7f and k e K . T h e n k = h~ u G / since H C\ U and u U. T h u s k E U, and we have A; G A T n /, and so u = h k G H ( K n t/), and the proof is complete.
l

The following shows how D e d e k i n d ' s l e m m a can be used. Before stating the result, however, we m e n t i o n t h a t intersections of a r b i t r a r y collections of subgroups are always subgroups. (This is clear since intersections of sugroups are closed under m u l t i p l i c a t i o n and inverses, and they contain the identity, so they cannot be empty.) X . 4 . C o r o l l a r y . Suppose t h a t H a n d K a r e s u b g r o u p s of a g r o u p G, a n d assume t h a t H K i s also a subgroup. L e t D = H n K , a n d let X a n d f V be t h e c o l l e c t i o n s of s u b g r o u p s d e f i n e d by X = { X|H C A C H K } and y = { Y\D C Y C K } .

D e f i n e t h e m a p 9 : X -> y by i n t e r s e c t i o n w i t h K , so t h a t 9 { X ) = X n K . T h e n 9 i s i n j e c t i v e , a n d i t s i m a g e i s t h e set W = { W G f V |W H = H W ) .

H K

Appendix:

The

Basics

329

In the diagram, the parallelogram formed by K , H K , H and D is a direct d i a m o n d , w h i c h means t h a t the lower node corresponds to the intersection of the subgroups represented by the side nodes, and the upper node corresponds to their product. T h e members of X lie along the upper edge, j o i n i n g H and H K , and the members of y lie along the parallel lower edge. T h e dashed lines represent the m a p 9, so that, for example, 9 { X ) = Y. In fact, each dashed line divides the o r i g i n a l figure into two new direct diamonds. F o r example, the upper parallelogram formed by the dashed line j o i n i n g X to Y is a direct d i a m o n d because X n K = Y and (as should be clear) X K = H K . T h e paralellogram formed by Y, X , H and D is also a direct d i a m o n d because H n Y = D and H Y = X , where as we shall see in the proof, the latter equality is a consequence of D e d e k i n d ' s l e m m a . T h e node on the lower edge t h a t is not joined to a corresponding node on the upper edge is intended to represent the fact t h a t the m a p 9 need not be surjective. P r o o f of C o r o l l a r y X . 4 . It should be clear t h a t intersection w i t h K actually does define a m a p from X to fV. N o w let X G X . T h e n H C X C H K , and therefore X = X n H K = H ( X n K ) = H 9 ( X ) , where the second equality follows by D e d e k i n d ' s l e m m a , w h i c h applies since H C X . T h i s shows t h a t X is determined by its image 9 { X ) , a n d thus 9 is injective. A l s o , since H 9 ( X ) = X is a subgroup, it follows by L e m m a X . l t h a t 9 ( X ) H = X 9 ( U ) , and thus 9 { X ) G W . F i n a l l y , i f W G W , then W H is a group by L e m m a X . l , and since W C K , we have W H G X . N o w 9 { W H ) = W t f n i f = W ( 7 f n K ) = W, where the second equality follows by D e d e k i n d ' s l e m m a and the t h i r d equality holds because H n K C W . Let G be an a r b i t r a r y group. If g G G , the subset G = | n G Z } is easily seen to be a subgroup, a n d clearly, G is contained i n every subgroup of G t h a t contains g . W e can thus describe G as the unique smallest subgroup of G t h a t contains g , where "smallest" should be understood i n the sense of containment. I n fact, if X is an any subset of G , there is a unique smallest subgroup t h a t contains X , namely the intersection of a l l of the subgroups of G t h a t contain X . ( T h i s makes sense because there certainly is at least one subgroup of G containing A , namely G itself, and i n general, intersections of subgroups are subgroups.) T h e unique smallest subgroup of G that contains X is said to be the subgroup generated by A , and it is denoted ( A ) . A m a x i m a l subgroup of a group G is a proper subgroup M such t h a t there is no subgroup H w i t h M < H < G. (It is clear t h a t every proper subgroup of a finite group is contained i n at least one m a x i m a l subgroup,

330

Appendix:

The

Basics

but some infinite groups, for example the a d d i t i v e group of the real numbers, have no m a x i m a l subgroups at all.) T h e intersection of a l l of the m a x i m a l subgroups of a finite group G is the F r a t t i n i subgroup, denoted $ ( G ) . A n interesting property of $ ( G ) is t h a t it is e x a c t l y the set of elements of G t h a t cannot contribute to the construction of a generating set for G. X.5. (A) L e m m a . L e t G be a < G, then finite group. If X C G i s a subset u $(G). such that

a l s o ( I U $ ( G ) } < G. Conversely, ( A ) < G, a l s o ( X U { u } ) < G, then

if u G G has the p r o p e r t y

t h a t whenever

P r o o f . L e t A C G be a subset such t h a t ( A ) < G . T h e n ( A ) is contained in some m a x i m a l subgroup M of G , a n d since also $ ( G ) C M , we see t h a t A U $ ( G ) C M , a n d thus ( A U $ ( G ) ) C M < G. Now suppose t h a t u G is an element such that ( X U { u } ) < G for a l l subsets A of G such t h a t ( X ) < G. T o prove that n e $ ( G ) , w e must show that u G M for a l l m a x i m a l subgroups M of G , so suppose u 0 M , where M is a m a x i m a l subgroup. T h e n ( M ) = M < G , and hence ( M u { u j ) < G , by hypothesis. W e have M < ( M U { } ) < G , a n d this is a c o n t r a d i c t i o n since M is a m a x i m a l subgroup. We

have seen that a subgroup G = (#) generated by a single element

l

g G G is e x a c t l y the set of powers of g i n G . ( A group generated by a single element is said to be cyclic.) If o { g ) is finite, the elements g are distinct for 0 < i < o ( g ) , a n d since these are easily seen to be a l l of the elements of G , it follows t h a t \ C \ = o { g ) i n this case. O n the other hand, if g has infinite order, it is clear t h a t G has infinite order. If G is a cyclic group, so t h a t G = (g) for some element g G G , it is easy to describe a l l of the subgroups of G . X.6. the L e m m a . L e t G be a c y c l i c g r o u p w i t h G = ( g ) , a n d l e t H C G be a g
m

n o n i d e n t i t y s u b g r o u p . Then smallest such

m

G H f o r some m d i v i d e s every

p o s i t i v e i n t e g e r m, a n d if m i s i n t e g e r n such that g
n

i n t e g e r , then

G H .

A l s o , H = ( g ) , a n d i n p a r t i c u l a r , a l l subgroups

of G a r e c y c l i c .

Proof. Since a l l elements of G are powers of g a n d H contains some noni d e n t i t y element, it follows t h a t g then g ~ form g
a a

G H for some nonzero integer a . If a < 0,

= (

)"

G H , a n d thus i n any case, H contains a n element of the

n

w i t h m > 0, a n d we can choose m to be as small as possible. G H for some integer n. B y the "division algon

Suppose now that g We g

m

r i t h m " we can write n = q m + r , where the "remainder" r satisfies 0 < r < m. have g
r

= g {g )- ,

a n d this element lies i n H since b o t h g

e H and

G H . Since r < m, it follows by the m i n i m a l i t y of m t h a t we cannot

have r > 0, and thus r = 0, a n d m divides n , as required.

Appendix:

The

Basics

331

To show t h a t H = ( g ) , we must show t h a t an a r b i t r a r y element h G H is a power of g . B u t G = ( g ) , so we c a n c e r t a i n l y w r i t e h = g for some integer n , and thus since g G H , we know t h a t n/m is an integer. T h e n h = g = { q ) l , a n d this completes the proof.
m m n n n m n m

X . 7 . C o r o l l a r y . L e t G = (g) be a finite c y c l i c g r o u p of o r d e r n , a n d f o r each p o s i t i v e d i v i s o r d o f n , l e t G = \ g ) . T h e n G i s the unique subgroup o f G h a v i n g o r d e r d, a n d these s u b g r o u p s G a r e a l l of t h e s u b g r o u p s ofG.

n / d d d d

P r o o f . Observe t h a t o ( g ) = \ G \ = n and \ G \ = o ( g ' ) . g = 1, a n d if 0 < e < d, then ( n / d ) e < n , and so { g ' ) t h a t o { g / ) = d, a n d thus \ G \ = d.
d n n d n d d

Now { g l ) = / 1. It follows

We show next t h a t an a r b i t r a r y subgroup H of G is one of the subgroups G. If H is t r i v i a l , it is G j , and so we can assume t h a t H is n o n t r i v i a l , and we let m be the smallest positive integer such t h a t g G H . Since g = l e H , L e m m a X . 6 guarantees that m divides n and t h a t H = { g } . It follows t h a t H = G , where d = n / m , and this completes the proof.
d m n m d

We have just seen t h a t if G is a finite cyclic group and H is a subgroup, then \ H \ divides | G | . I n fact, this is true for a l l finite groups G, even if they are not cyclic. T o establish this theorem of Lagrange, we consider cosets. G i v e n a subgroup H C G and a n element x G G , where G is a n a r b i t r a r y group, the set H x { h x \ h G H } is a r i g h t c o s e t of H i n G . T h e number of these right cosets, w h i c h m a y be infinite, is the i n d e x of H i n G , w h i c h is denoted | G : H \ . ( O f course, the index is the number of d i f f e r e n t right cosets of H i n G . If H x = H y , then this coset w i t h two names should be counted only once.) X . 8 . T h e o r e m (Lagrange). L e t H be a s u b g r o u p of a n a r b i t r a r y g r o u p The f o l l o w i n g then hold. (a) L e t y G H x , w h e r e x G G . T h e n H y = H x . (b) D i s t i n c t r i g h t cosets (c) A l l r i g h t cosets (d) I f \ G \ i s finite, of H i n G a r e disjoint. G.

of H i n G h a v e c a r d i n a l i t y e q u a l t o \ H \ . then \ H \ divides \G\ a n d \ G \ / \ H \ = \ G : H \ .

P r o o f . W e show first that i f h G H , t h e n H h = H . (Note t h a t this is the special case of (a) where x = 1 a n d y = h . ) Since h ' G H and H is closed under m u l t i p l i c a t i o n , we have H h ~ C H . R i g h t m u l t i p l i c a t i o n by h yields H C H h , and the reverse containment is another consequence of the fact t h a t H is closed under m u l t i p l i c a t i o n . A s s e r t i o n (a) follows since we can w r i t e y = hx, where h G H , a n d thus H y = H h x = H x . N o w (b) is immediate since i f z G H x n H y , then H x = H z = H y by two applications of (a).
1 l

332

Appendix:

The

Basics

For (c), we construct a bijection 9 : H - > H x , where Ex is an a r b i t r a r y right coset of i f i n G . L e t 9 ( h ) = hx and observe t h a t 9 is a u t o m a t i c a l l y surjective since by definition, every element of H x has the form hx for some element he H . A l s o , 9 is injective since i f hx = kx for elements h , k G H , t h e n right m u l t i p l i c a t i o n by x ~ yields h = k.
l

F i n a l l y , since # G i f p for a l l g G G , we see t h a t G is the u n i o n of the | G : i f | right cosets of H i n G . These cosets are disjoint by (b), and each of t h e m has c a r d i n a l i t y | f f | by (c), a n d thus | G | = | G : / f | | / f | , p r o v i n g (d). X.9. \G\. C o r o l l a r y . L e t g G G, w h e r e G i s a finite group. T h e n o(g) divides

P r o o f . W e have o { g ) = \ ( g ) \ , w h i c h divides | G | by Lagrange's theorem.

T h e exponent of a finite group is the least c o m m o n m u l t i p l e of the orders of its elements. B y C o r o l l a r y X . 9 , therefore, the exponent of G divides \ G \ for finite groups G . E v e r y t h i n g we d i d w i t h right cosets works as well w i t h left cosets of the form x H = { x h \ h G i f } . I n particular, the "mirror image" of Theor e m X . 8 ( d ) shows that if i f C G , where G is finite, then the number of left cosets of H i n G is also equal to | G | / | f f |, and thus the sets { H x \ x G G } and { x H | x G G } have equal c a r d i n a l i t y (although i n general, they are different sets). B u t even if G is infinite, the cardinalities of the sets of left and right cosets of i f i n G are equal. T o see this, consider the bijection from the set of a l l subsets of G to itself defined by A ^ { x ~ \ x G A } , and check that it carries each left coset of i f i n G to a right coset and v i c e v e r s a .
l

X.10. C o r o l l a r y . L e t H C AT C G , w h e r e G i s a K are subgroups. Then \ G: i f | = |G : K \ \ K : H \ .

finite

group and H

and

P r o o f . Since | G : K \ = \ G \ / \ K \ and \ K : H | = \ K \ / \ H | , we conclude that | G : K \ \ K : H \ = \ G \ / \ H | = | G : i f |. Then

X.ll. C o r o l l a r y . L e t i f a n d K be s u b g r o u p s of a finite g r o u p G. \ K : H n K \ < \ G : H \ , a n d e q u a l i t y h o l d s if a n d o n l y if H K = G. P r o o f . Since H K is a subset of G , L e m m a X . 2 yields

IG| >
1 1

\HK\ =
1 1

l^l^l
|ff n K \ '

w i t h equality i f and only if i f AT = G . T h e n

again w i t h equality if a n d only if i f K

= G.

Appendix:

T h e Basics

333

A n o t h e r useful fact i n this vein is the following. X . 1 2 . C o r o l l a r y . L e t H a n d K be s u b g r o u p s pose of a finite g r o u p G, a n d s u p -

that \ G :H \ a n d \ G :K \ are relatively p r i m e .

T h e n H K = G.

P r o o f . Since i i i l i f C i i C G , it follows by C o r o l l a r y X . 1 0 t h a t \ G : H \ divides \ G : H n K \ , and similarly, \ G : K \ divides \ G : H D K \ . Since the indices \ G : H \ and \ G : K \ are coprime, their product must divide \ G : H n K \ , and thus

T h i s yields \G:H\ = ^
' '

\ < , , = IA/ : P D P I , |P| ~ \ H H K \ '

1 1

and thus by C o r o l l a r y X . l l , equality holds and H K = G.

Versions of these corollaries hold even for infinite groups but we w i l l not give proofs, or even precise statements. a n a Gr are groups, a Dijection y : b i -> Gr is said to be an i s o m o r p m s m it respects m u l t i p l i c a t i o n , by w n i c n we m e a n t n a t ( ^ y j V K x m v ) r an elements x , y e L a . A l s o it tnere is an i s o m o r p h i s m i r o m
2 2

r
,
,

~ r rrt
~
f

It
1 8

iTwh i T o r
* '
S m

a r e

*
~

s o m o r p m c

'
l S

a i

?.

w e

Oi,

J;
"'

"
1 8

^ is an i s o m o r p n i s m ^ t h e n
1 1 U l t n e n a O

a n a m iact,

is an equivalence relation.j

' .

^2 -

A group theorist w o u l d say t h a t a group is n o t h i n g but a certain m u l t i p l i c a t i o n rule on a set, and t h a t what counts is the m u l t i p l i c a t i o n rule, and not the set. F r o m this point of view, we see t h a t i f the u n d e r l y i n g set of a group is changed, but the m u l t i p l i c a t i o n is somehow preserved, this w i l l not fundamentally alter the group. T o a group theorist, therefore, isomorphic groups are essentially identical, and so if G =i G , then G enjoys a l l of the "group theoretic" properties of G . F o r example, if G has exactly eight elements of order 3, then the same is true of G , a n d i f G\ has an abelian subgroup of order 10, then so too does G . ( A group G is abelian if xy = yx for a l l x,y < G G.)
x 2 2 x x 2 2

O f course, isomorphic groups m a y look very different to someone w h o is not a group theorist. F o r example, the additive group of the real numbers is isomorphic to the m u l t i p l i c a t i v e group of the positive real numbers v i a the m a p x ^ e . M o r e surprisingly, the additive group of the real numbers is isomorphic to the additive group of the complex numbers. (It requires an appeal to the a x i o m of choice to establish the existence of such an isomorp h i s m , however, and so it is not possible to describe an explicit map.)
x

334

Appendix:

The

Basics

The following l e m m a shows t h a t for each positive integer n , there is essentially only one cyclic group of order n . X . 1 3 . L e m m a . L e t B = (b) a n d C = (c) be c y c l i c g r o u p s of t h e same f i n i t e o r d e r n , a n d l e t 9 : B - > C be d e f i n e d by 9 ( b ) = c f o r a l l i n t e g e r s i. T h e n 9 i s a w e l l d e f i n e d i s o m o r p h i s m f r o m B o n t o C.
l %

P r o o f . F i r s t , observe t h a t o ( b ) = n = o(c). If b = V, then i = j m o d n , and thus & = c\ and it follows t h a t B is unambiguously defined. A l s o , 9 is surjective since every element of C has the form c? for some integer j , and since \ B \ = \ C \ < oo, it follows t h a t 9 is also injective. F i n a l l y , i f x,y E B , write x = b and y = b for integers r and s. T h e n
r s

9(b b ) = 0(b and

r + s

) = c

r + c

= cc

r s

= 9(b )0(b ),

so 9 ( x y ) = 6 ( x ) 6 ( y ) , and 0 is an i s o m o r p h i s m .

If 9 : G i -> G is a n isomorphism, it should be clear t h a t 9 carries each subgroup of G i to a subgroup of G , and i n fact, the m a p H ^ 0 ( H ) is a bijection from the set of subgroups of G onto the set of subgroups of G . Furthermore, i f H C G is a subgroup that satisfies a certain property i n G , then its image 0 ( H ) satisfies the corresponding property i n G .
2 2 x 2 x x 2

G i v e n a group G , consider, for example, the set of elements z G G that commute w i t h a l l elements of G . T h i s set, denoted Z ( G ) , is the c e n t e r of G , and it is easy to check that it is a subgroup. If 0 : G - > G is an isomorphism, it should be clear t h a t 9 carries the center of G to the center of G . M o r e generally, if H is any subgroup of G that can be described w i t h the definite article "the", then 9 ( H ) is the corresponding subgroup of G . For example, if H is the F r a t t i n i subgroup of G or the derived subgroup of G i , then the image 9 ( H ) is respectively the F r a t t i n i subgroup or the derived subgroup of G . (It is not necessary to k n o w the definition of the derived subgroup to see t h a t this must be true.)
x 2 x 2 x 2 x 2

A n i s o m o r p h i s m from a group G to itself is called an a u t o m o r p h i s m of G , and the set of a l l automorphisms of G is d e n o t e d - A u t ( G ) . O f course, A u t ( G ) is a subset of S y m ( G ) , and it is easy to see t h a t i n fact, A u t ( G ) is a subgroup of S y m ( G ) . X.14. L e m m a . L e t G be a c y c l i c g r o u p . T h e n A u t ( G ) i s a b e l i a n , a n d if G i s finite, t h e n | A u t ( G ) | = y>(|G|), w h e r e ip i s E u l e r ' s t o t i e n t f u n c t i o n . P r o o f . W r i t e G = ( g ) . Observe first t h a t if a , (5 A u t ( G ) and a ( g ) = ( 5 ( g ) , then a ( g ) = a ( g ) = ( 5 ( g ) = f 3 ( g ) for a l l integers n . Since every element of G has the form g for some integer n , it follows t h a t a and (5 agree o n a l l of G , and thus a = (3.
n n n n n

Appendix:

The Basics

335

Now let o~, r G A u t ( G ) . T h e n <r(g) = g a n d r ( g ) = g for some integers s a n d t, and it is easy to check t h a t <JT a n d T O b o t h carry g to g . Since CTT and T O agree on we conclude t h a t these automorphisms are equal, and hence A u t ( G ) is abelian.
s t s t

Now suppose \ G \ = n , so t h a t o ( g ) = n . Since automorphisms preserve orders of elements, every a u t o m o r p h i s m of G maps g to a n element of the set X = { x G G | o (x) = n}, a n d by the first paragraph of the proof, distinct automorphisms of G carry g to distinct elements of X . If x G A is arbitrary, then G = ( x ) , and hence by L e m m a X . 1 3 , there is an i s o m o r p h i s m 9 from G = (g) to G = (x) such that % ) = x. T h u s 9 is an a u t o m o r p h i s m of G t h a t carries # to x, and it follows t h a t | A u t ( G ) | = | A | . To complete the proof, we argue t h a t | A | = <p(n). Since every element of G is uniquely of the form g w i t h 0 < r < n , it suffices to show that g G A if and only if r and n are coprime. T o see this, let i f = ( g ) , so that \ H \ = o ( g ) , a n d let m be the unique smallest positive integer such that g G i f . It follows by L e m m a X . 6 , t h a t m divides b o t h r and n, and thus if r and n are coprime, we have m = 1 and g G i f , and thus H = G. T h e n ( ) = | i f | = , and # G X .
r r r r m r r 0 5 n

Conversely, suppose that g G A and let m be the greatest c o m m o n divisor of r and n. W r i t e d = n / m , so t h a t n = md, w h i c h divides r d , and thus ( ) = 1 and d > o ( g ) = n . T h e n d = n and m = 1, as wanted.
r r d r 5

A subgroup H C G is characteristic if every a u t o m o r p h i s m of G maps i f onto i f . Since a n isomorphism from one group to another carries the center to the center and the F r a t t i n i subgroup to the F r a t t i n i subgroup, it follows t h a t an a u t o m o r p h i s m of a group G carries Z ( G ) to Z ( G ) and $ ( G ) to $ ( G ) . M o r e generally, every subgroup of a group G t h a t c a n be described unambiguously using the w o r d "the" must be m a p p e d to itself by all automorphisms of G , and hence such a subgroup is characteristic. In particular, if G is a finite cyclic group, then every subgroup of G is characteristic because if i f C G and | i f | = d, then by C o r o l l a r y X . 7 , we can say t h a t " i f is t h e subgroup of G h a v i n g order d " . G i v e n elements x and g i n a group G , we write x to denote the element g ~ x g , and we say that x is conjugate to x i n G . (Note t h a t x = x if and o n l y i f x g = g x . ) It is easy to see t h a t conjugation by g is a n a u t o m o r p h i s m of G . (This m a p is b o t h injective and surjective because it has an inverse, namely conjugation by g ~ \ a n d it is t r i v i a l to check t h a t {xy)<> = x y as required.) T h e a u t o m o r p h i s m x ^ x is said to be an inner a u t o m o r p h i s m of G , and the set of a l l inner automorphisms of G is denoted I n n ( G ) . Since ( x ) = x , for a l l x , g , h G G , it is easy to see t h a t I n n ( G ) is a subgroup of the full a u t o m o r p h i s m group A u t ( G ) .
l 9 9 9 9 9 9 h 9 h

336

Appendix:

The

Basics

If X C G is a subset and g G, then the image of X under conjugation by g is the subset X = { x \ x G } . Since the conjugation maps are automorphisms, it follows that if H C G is a subgroup, then a l l of its conjugates H for g G G are subgroups too. N o w recall that a characteristic subgroup of G is a subgroup that is m a p p e d to itself by a l l automorphisms of G . W e can weaken this c o n d i t i o n and consider the n o r m a l subgroups of G , w h i c h are those subgroups t h a t are m a p p e d to themselves by a l l i n n e r automorphisms of G . A subgroup H C G is n o r m a l , therefore, precisely w h e n H = H for a l l elements g G G . W e write H < G i n this case. O f course, characteristic subgroups are a u t o m a t i c a l l y n o r m a l .
9 9 9 9

In general, it is not true t h a t n o r m a l subgroups of n o r m a l subgroups of a group G are n o r m a l i n G , but the following substitute is often useful. X . 1 5 . L e m m a . L e t N < G, w h e r e G i s a g r o u p , a n d l e t C be c h a r a c t e r i s t i c i n N . T h e n C < G. P r o o f . L e t g G G and let a be the inner a u t o m o r p h i s m of G defined by conjugation by g . T h e n a is injective, and it maps N onto N , a n d since a respects m u l t i p l i c a t i o n , the restriction of a to A is an a u t o m o r p h i s m of N . A s C is characteristic i n N , we have C = a { C ) = C, and thus C < G. U
9

G i v e n a subgroup H C G , the set {# G G | H = H } , is the normalizer of i n G , w h i c h is denoted N ( t f ) . It is easy to see t h a t N { H ) is always a subgroup of G , a n d that it contains the subgroup H . It follows t h a t N ( J f ) is the unique largest subgroup of G that contains H as a n o r m a l subgroup.
9 G G G

Since conjugation by an element g G G is an injective map, \ H \ = \ H \ for subgroups H of G . T h u s if \ H \ is finite and Jf C H , then Jf = H , and 5 G N ( J f ) . I n general, however, it can happen that H is proper i n H . B u t if H C H for elements g G G , then also i f " C P . C o n j u g a t i n g b o t h sides by g , we get H C and thus i n fact, tf = i f . T o show t h a t H < G , therefore, it suffices to check that H <Z H for a l l elements g G G .
9 9 G 9 3 1 5 9

X . 1 6 . L e m m a . A e i i f k a s u b g r o u p of a g r o u p G. T h e n t h e f o l l o w i n g a r e equivalent. (1) H < G. (2) H x = x H f o r a l l x G G . (3) E v e n / h</fti cose* of H i n G i s a left (4) E v e n / ie/t coset (5) ( H x ) ( H y ) coset of H . of H .

of H i n G i s a r i g h t coset x , y e G .

= H x y for all

(6) T h e set of r i g h t cosets

of H i n G i s c l o s e d u n d e r m u l t i p l i c a t i o n .

Appendix:

The

Basics

337

P r o o f . It is clear t h a t (2) implies (4). Conversely, i f (4) holds and i e G , t h e n the left coset x E is a right coset of H c o n t a i n i n g x , and so it must be the coset Ex by T h e o r e m X . 8 ( a ) , and this proves (2). T h u s (2) a n d (4) are equivalent, a n d since (2) is left-right s y m m e t r i c , it must also be equivalent to (3), w h i c h is the m i r r o r image of (4). A l s o , (5) and (6) are easily seen to be equivalent. T h a t (5) implies (6) is obvious, a n d conversely, assuming t h a t E x H y is a right coset of H , t h e n since it contains the element xy, it must be the right coset Exy by T h e o r e m X . 8 ( a ) . Now assume (1), so t h a t E < G, a n d let x G G . T h e n E = H = x ~ E x , and left m u l t i p l i c a t i o n by x yields x H = H x , and thus (2) holds. N o w assume (2), a n d let x,y G G. T h e n H x H y = H ( x E ) y = E ( H x ) y = Exy since E E = E , and this proves (5). T o complete the proof, it suffices to show t h a t (5) implies (1). T o see t h i s , assume (5) a n d let x G G. T h e n
x x

= x~ Hx C E x '

Ex

= Ex~ x = E ,

a n d since this holds for all x G G, we have H < G.

A useful observation about a n o r m a l subgroup N < G is that i f H C G is an a r b i t r a r y subgroup, then N E is a subgroup. T o see w h y this is so, observe t h a t N E = a n d apply L e m m a X . l . Now let A be a n o r m a l subgroup of G, where G is an a r b i t r a r y group, and w r i t e G / N to denote the set of right cosets of N i n G. (Of course, L e m m a X . 1 6 implies t h a t the w o r d "right" i n the previous sentence is superfluous because for n o r m a l subgroups, there is no d i s t i n c t i o n between left cosets a n d right cosets.) B y c o n d i t i o n (5) of L e m m a X . 1 6 , the set G / N is closed under m u l t i p l i c a t i o n , and we observe t h a t i n fact it is a group since the coset N l = N acts as a n identity and ( N x ) ( N - ) = N l , so t h a t A ^ x " is the inverse of N x i n G/N. T h e group G / N is called the quotient g r o u p or factor g r o u p of G w i t h respect to the n o r m a l subgroup N . (We stress that N must be n o r m a l ; otherwise by L e m m a X . 1 6 , there are right cosets whose product is not a right coset.)
1 1 X

[j
heH

N h=

h N=

E N ,

h e H

If G a n d H are a r b i t r a r y groups, a m a p 9 : G - > H is called a h o m o m o r p h i s m if 9 ( x y ) = 9 ( x ) 9 ( y ) for a l l x,y G G. (Thus an i s o m o r p h i s m is a h o m o m o r p h i s m t h a t happens to be b o t h injective a n d surjective.) We consider some examples of h o m o m o r p h i s m s . G i v e n an a r b i t r a r y group G , let 9 { x ) G A u t ( G ) be conjugation by x . It is t r i v i a l to check that g y = ( g ) y for elements g , x , y G G , and thus 9 { x y ) = 9 ( x ) 9 ( y ) a n d 9 : G -> A u t ( G ) is a h o m o m o r p h i s m . (Remember t h a t m u l t i p l i c a t i o n i n A u t ( G ) is defined to be left-to-right.) T h e image of G under 9, of course,
x x

338

Appendix:

T h e Basics

is I n n ( G ) . For another example, let Z be the additive group of the integers and let H be a n a r b i t r a r y group. F i x an element h G H and consider the m a p n i-> h from Z to H . Since h = h h for m, n G Z , this m a p is a h o m o m o r p h i s m , and the image of Z i n J i is clearly the subgroup ( h ) .
n m + n m n

Perhaps the most i m p o r t a n t example is the c a n o n i c a l h o m o m o r p h i s m 7T : G -> G / A , where N < G and 7r(s) = A ^ . Observe that TT really is a h o m o m o r p h i s m since vr(xy) = W x y = ( N x ) ( N y ) = 7r(x)7r(y) for elements x , y G G . T h e canonical h o m o m o r p h i s m TT is always surjective, but since all elements of N m a p to the identity of G/N, it is never injective unless N = 1, the t r i v i a l subgroup. G i v e n an a r b i t r a r y group h o m o m o r p h i s m 0 : G -> Jf, the k e r n e l of 0, denoted is the set { x G G | 0(x) = 1}. For example, the kernel of the h o m o m o r p h i s m 0 : G -> A u t ( G ) , where 0(x) is conjugation by x , is the set of elements x G G such that g = g for all g G G . Since # = 5 i f and only if gx = x g , we see that = Z ( G ) , the center of G . F o r another example, consider the canonical h o m o m o r p h i s m TT : G -> G / W , where N < G. Since the identity of G / A is the coset N , we see that g lies i n ker(vr) precisely w h e n N g = N , and thus ker(vr) = A".

ker(0)

ker(0)

X.17.

L e m m a . L e t 9 : G - H be a g r o u p h o m o m o r p h i s m ,

a n d let N

ker(0). T h e f o l l o w i n g t h e n h o l d . (a) 9 ( 1 ) =
l

1.

(b) 9 ( x - ) = 9 ( x ) -

f o r x e G .

(c) N i s a n o r m a l s u b g r o u p o f G . (d) F o r x , y G G, w e h a v e 9 ( x ) = 9 ( y ) if a n d o n l y if N x = (e) 9 i s i n j e c t i v e if a n d o n l y if N = 1. P r o o f . W e have = = 0(1), a n d left m u l t i p l i c a t i o n by 0(1)~ i n H yields 0(1) = 1, p r o v i n g (a). A l s o , if x G G , we have 1 = 0(1) = 1 0(xx" : ) = 0(x)0(x" ), a n d (b) follows by left m u l t i p l i c a t i o n by 9 ( x ) ~ .
1
x

Ny.

0(1)0(1) 0(1-1)

Now N is nonempty since 1 G AT by (a). 9 ( x ) 9 ( y ) = 1-1 = 1, and thus xy G N , and N : Furthermore, if x G N , then 0(x" ) = 9 ( x ) ~ and this proves that A is a subgroup. T o n G N and g G G . T h e n 9(n) = 9 ( g - n g ) n G N . T h i s shows t h a t N C N for a l l g (c).
l l 9 g

A l s o , if x, y G N , then 9 ( x y ) = is closed under m u l t i p l i c a t i o n . = l " = and thus x " G N , complete the proof of (c), let = 9 ( g ) - 1 9 ( g ) = 1, and hence G G , and thus N < G, p r o v i n g
1

1,

For (d), observe t h a t 0(x)0(y)" = 0(xy" ). T h e left side is equal to if and only if 0(x) = 9 ( y ) , and the right side is 1 if and only if x y " G N , or e q u i v a l e n t s , x G Ny, and this, we know, is equivalent to N x = Ny. T h i s proves (d), and (e) is a n immediate consequence.
1
1

Appendix:

The

Basics

339

T h u s kernels of homomorphisms denned on a group G are n o r m a l subgroups. Since we saw previously t h a t every n o r m a l subgroup N < G is the kernel of the canonical h o m o m o r p h i s m G > G/N, it follows that the n o r m a l subgroups of G are exactly the kernels of the h o m o m o r p h i s m s defined on G. T h e following h o m o m o r p h i s m theorem asserts that a n a r b i t r a r y surjective h o m o r p h i s m 6 is essentially the canonical h o m o m o r p h i s m from G onto G/N, where N = ker (0).

X . 1 8 . T h e o r e m ( H o m o m o r p h i s m ) . L e t 6 : G - > H be a s u r j e c t i v e g r o u p h o m o m o r p h i s m , a n d l e t N = ker (0). Then G / N ^ H . I nfact, there i sa n i s o m o r p h i s m r : G / N - > H such t h a t T ( N X ) = 9 { x ) f o r a l l x e G . P r o o f . L e t I G G . B y L e m m a X . 17(d), a l l elements of the coset TVx have the same image under 6, and this image is 6 { x ) . W e can thus unambiguously define r on G / N by setting r ( N x ) = 9 { x ) , a n d we must show that r is an isomorphism. First, r{NxNy) = r ( N x y ) = 6{xy) = 9{x)6{y) = T ( N x ) r ( N y ) ,

a n d thus r is a h o m o m o r p h i s m . A l s o , since r ( N x ) = 9 ( x ) , we see t h a t the image of r is the image of 6, and since this is a l l of H , it follows that r is surjective. T o prove that r is injective, it suffices to check that k e r ( r ) is t r i v i a l , and so we suppose that Nx e k e r ( r ) . T h e n 1 = T { N X ) = 9 ( x ) , and thus x ker(0) = N , and so Nx = N , w h i c h is the identity of G/N. U A c t u a l l y , T h e o r e m X . 1 8 even tells us something about homomorphisms that are not surjective. T o e x p l a i n this, suppose that 6 : G H is an a r b i t r a r y h o m o m o r p h i s m . If A C G is a subgroup, it is easy to check that the image 6 { X ) is a subgroup of H , and i n p a r t i c u l a r , 0 { G ) is a subgroup of H . W e can thus view 9 as a surjective h o m o m o r p h i s m from G onto 6 { G ) , a n d so we can a p p l y the h o m o m o r p h i s m theorem to deduce that 6 ( G ) ^ G /ker (0). A s a n example of how the h o m o m o r p h i s m theorem can be used, we prove the following so-called " N / C - T h e o r e m " . T o state it, we define the centralizer i n G of a subset A C G to be the set of elements of G that commute w i t h a l l elements of A . It is easy to check that this set, denoted C ( A ) , is a c t u a l l y a subgroup of G .
G

X . 1 9 . C o r o l l a r y . L e t H be a s u b g r o u p of a n a r b i t r a r y g r o u p G, a n d w r i t e N = N (U) a n dC = C { H ) . Then C < N a n dN / C is isomorphic t o a subgroup o f A u t ( H ) .

G G

P r o o f , l i x N , then H = H , and we w r i t e r : H - > H to denote the m a p h ^ h . Since conjugation by x is a n a u t o m o r p h i s m of G , it is easy to see t h a t T is an a u t o m o r p h i s m of H . A l s o , if x , y G N and h e H , t h e n (h)T T = (h )y= h = [ h ) T , a n d thus r = T T , a n d it follows t h a t
x x X x x y x y X Y x y X V

340

Appendix:

The

Basics

the m a p 0 : N element x H equivalents, h thus k e v ( 0 ) = C complete.

-> A u t ( i f ) denned by 0 ( x ) = T is a h o m o m o r p h i s m . A n lies i n ker(0) if and only i f r is the identity m a p on i f , or = h for a l l h e H . B u t / i = h precisely when x h = hx, and ( H ) = C. T h e n N / C 6 ( N ) C A u t f i f ) , and the proof is
X x x

If i f is a cyclic subgroup of G , for example, then N ( H ) / C ( H ) is a n abelian group since it is isomorphic to a subgroup of A u t ( i f ) , w h i c h is abelian by L e m m a X . 1 4 .
G G

A n o t h e r a p p l i c a t i o n of the h o m o m o r p h i s m theorem is the following. X.20. C o r o l l a r y . L e t N a n d H be s u b g r o u p s of a g r o u p G, a n d suppose

t h a t N < G. T h e n H H N < H a n d N H / N

H / ( H n N).

P r o o f . R e c a l l t h a t N H is a group, and observe that the right cosets of A" in W i f a l l have the form N n h = N h , where n G N and h G i f . N o w define 6 : i f -> N H / N by setting 0 ( h ) = N h , and observe that 6 is surjective, and that it is the restriction to i f of the canonical h o m o m o r p h i s m TT : G - G /iV. It follows t h a t 0 is a h o m o m o r p h i s m , and k e r ( 6 ) = H D k e r ( n ) = H H N . T h u s H n N < H and i f / ( i f n J V ) ^ A ^ i f / A by the h o m o m o r p h i s m theorem. We have already mentioned that if 0 : G - > H is an a r b i t r a r y group h o m o m o r p h i s m and X C G is a subgroup, then 0 ( A ) is a subgroup of i f . The following correspondence theorem shows that if 0 is surjective, then every subgroup of H is the image of some subgroup of G . In fact, much more is true. X.21. T h e o r e m (Correspondence). L e t 0 : G -> H h o m o m o r p h i s m , a n d let N = kex(0). T h e n the m a p X f r o m t h e set X of s u b g r o u p s of G t h a t c o n t a i n N s u b g r o u p s of H . F u r t h e r m o r e , if X G X a n d Y G y the f o l l o w i n g hold. (a) A = { x G G | 0 ( x ) G Y } . (b) T h e m a p a d e f i n e d by X g H-> 0 ( X g ) i s a b i j e c t i o n f r o m t h e set of a l l r i g h t cosets of X i n G o n t o t h e set of a l l r i g h t cosets of Y i n H , a n d i n p a r t i c u l a r , \ G : A | = | i f : Y\. ( c ) X < G if a n d o n l y i f Y < i f , a n d i n t h i s case, an isomorphism G / X 9* H / Y . t h e m a p a of (b) i s be a s u r j e c t i v e g r o u p H- 0 ( A ) i s a b i j e c t i o n o n t o t h e set y of a l l a n d Y = 0(A), then

P r o o f . T o prove that the m a p A H- 0 ( A ) is a bijection from X onto y , it suffices to construct an inverse m a p from y to X . G i v e n a subgroup Y C i f , write 0~ (Y) = { x G G | 0 ( x ) G Y } , so t h a t 0~ (Y) is the inverse image of
l l

Appendix:

The

Basics

341

Y i n G . ( C a u t i o n : we are not asserting t h a t there is a m a p called 0 " . ) It is easy to check t h a t 9 ~ ( Y ) is a subgroup of G, and since it clearly contains ker(0) = N , we have 9 ~ { Y ) G X .
1 1 1

We

proceed now to show t h a t the m a p Y ^ 9 { X ) from X to y.

1

_ 1

( Y ) from y to X is the T o prove equality, let

inverse of the m a p X ^

( A s s e r t i o n (a) w i l l also: follow C Y.

from this.) If Y G y, then clearly 9 ( 0 - ( Y ) )

1 r

y G Y and observe t h a t since 9 is surjective, there exists a n elenent x G G such t h a t 0(x) = y. T h e n x G 0 " ( * ) , and thus y = 9 ( x ) lies i n 0 ( 0 - ! ( Y ) ) , a n d we have 0 ( 0 ( * ) ) = Y, as needed.
_ 1 r

If A G A ' , then clearly A C 0 " ( 0 ( A ) ) . T o prove equality, we consider an element g G 0 " ( 0 ( A ) ) , a n d we show t h a t g G A . C e r t a i n l y , 0(#) G 0(A), and thus there exists x G A such t h a t 9 ( g ) = 0 ( x ) . It follows by L e m m a X . 1 7 ( d ) t h a t N g = N x , a n d thus g G N g = N x C A , where the final containment holds because J V C I b y the definition of X . T h u s A = 0 " ( 0 ( A ) , and this shows t h a t the maps A ^ 0 ( A ) and Y i-> 0 - ( Y ) are inverse bijections between # and y, as required.
1 X :

Now l e t X E X a n d F e ^ , where y = 0 ( A ) . If g G G, then 0 ( A # ) = 9 ( X ) 9 ( g ) = Y 9 { g ) , a n d so a really is a m a p from the set { X g \ g e G } to the set { Y h | ft G # } . A l s o , a is surjective since if h G H , then ft = 0(g) for some element g G G, and thus a ( A y ) = 9 { X g ) = Y f t . T o prove t h a t a is injective, suppose t h a t a ( A a ) = a ( A 6 ) for elements a, 6 G G . T h e n Y 0 ( a ) = Y 0 ( 6 ) , a n d so 0(6) = y 0 { o ) for some element y G Y. Since Y = 0 ( A ) , we have y = 0(x) for some element x G A , a n d thus 0(6) = 0(x)0(a) = 0(xa). It follows t h a t A^6 = A ^ x a , a n d since N C. X , this coset of N is contained i n b o t h X a and A 6 , w h i c h , therefore, are not disjoint. It follows that X a = X b , a n d thus a is injectve, and the proof of (b) is complete. Observe that 9 ( X )
9

= 0 ( A ) ^ ) = Y ^)

for

a l l elements g G G. N o w

suppose t h a t A < G. L e t ft G H , and write ft = 9 { g ) , where g G G . T h e n = y ( s ) = 9 ( X 9 ) = 0 ( A ) = Y , a n d thus Y < H . Conversely, suppose t h a t Y < H . L e t g G G , a n d observe t h a t N = N<> C X*>, so that A G A l s o , 0 ( A 9 ) = Y ( f ) = Y = 0 ( A ) , a n d thus A = A since the m a p defined on X by a p p l y i n g 0 is injective. It follows t h a t A < G , a n d this shows t h a t A < G if and o n l y if Y < i f , as required.
9 0 9

F i n a l l y , assuming t h a t A < G and Y < we must show that the m a p a : G / X > H / Y is an isomorphism. W e k n o w from (b) that a is b o t h injective a n d surjective, and so it suffices to show t h a t it is a h o m o m o r p h i s m . T h i s is clear, however, since i f a, b G G , then 9 { X a X b ) = 0 ( A a ) 0 ( A 6 ) .

If A G X i n the n o t a t i o n of the correspondence theorem, then clearly N (A) G X , and we argue t h a t 0 ( N ( A ) ) = N { 9 { X ) ) . T o see w h y this is true, observe t h a t since the normalizer of a subgroup is the largest
G G H

342

Appendix:

The

Basics

subgroup i n w h i c h it is n o r m a l , it suffices to show t h a t i f U, V G X w i t h U Q V , t h e n U < V if and o n l y i f 9 ( U ) < 6 { V ) . T h i s follows by a p p l y i n g the correspondence theorem to the surjective h o m o m o r p h i s m from V to 9 ( V ) obtained by restricting 9 to V. N e x t , we apply the correspondence theorem to the canonical homomorp h i s m 7T : G > G/N, where N < G. T h e n ker (7r) = N , so X is the set of a l l subgroups of G t h a t contain N . If X G X , then T T ( X ) = X / N , and thus the correspondence theorem tells us t h a t every subgroup of G / N is uniquely of the form X / N , where X is a subgroup satisfying TV C X C G. A l s o , A < G if a n d o n l y i f X / N < G/N, and i n t h a t case, G / X (G/N)/(X/N). We complete our review by discussing direct products. Suppose t h a t we are given (not necessarily distinct) groups d w i t h 1 < i < r . L e t P be the set of ordered r-tuples ( x , x , . . . , x ) , and define m u l t i p l i c a t i o n i n P componentwise, so that
x 2 n

(xi,x , -,x )(yi,y ,.. .,y ) = (x y ,x y ,..

2 r 2 r x x 2 2

.,x y )
r r

It is t r i v i a l to check t h a t P is a group; it is called the external direct p r o d u c t of the groups G , and we write P = G x G x x G .
{ x 2 r

If 1 < i < r , let N i C P be the set of r-tuples of the form (1,...,1,^,1,...,1), where the component i n p o s i t i o n i is an a r b i t r a r y element G G and all other components are the identities of the respective groups. It is easy to see t h a t N i is a subgroup of P , a n d that N i = G i . A l s o , it should be clear that every element x G P is uniquely of the form x = n n - - - n w i t h r n N and i n particular, we can write P = N N - N .
X i t x 2 r i t X 2 r

The elements of N i a n d N j commute i i i ^ j , and so if 1 < k < r , we have N i C C ( N ) C N ( A ^ ) for i ^ k. Since A^ is also contained i n N p ( A T ) , w h i c h is a subgroup, we see t h a t P = N N N C N ( N ) , and thus N < P .
P k P f e fe fe X 2 r P k k

G i v e n a group G a n d subgroups < G for 1 < i < r , we say t h a t G is the internal direct p r o d u c t of the n o r m a l subgroups M provided t h a t every element g G G is uniquely of the form = m m m with m , G M , . T h u s if P is the external direct product of the abstract groups G, then P is also the internal direct product of the n o r m a l subgroups N i t h a t we defined above, w i t h N i ^ d .
5 x 2 r

N o t e t h a t "external direct product" is a c o n s t r u c t i o n since given an a r b i t r a r y finite list of groups, one can define their external direct product. B u t "internal direct product" is really a statement about a finite collection of n o r m a l subgroups of a given group; it may be true, but it is not a u t o m a t i c a l l y so. Despite these differences, and at the risk of confusion, we often use the

Appendix:

The

Basics

343

same notation, and we write G = M direct p r o d u c t of its subgroups M .

x

x M

x x M

if G is the internal

G i v e n a group G and subgroups M , < G for 1 < i < r , suppose that G = M \ M M . T h e following result shows what a d d i t i o n a l information is needed to prove t h a t G = M x M x x M .
2 r x 2 r

X.22. if

Theorem. Let G = M

M , where G is a group and M

r x

is a only

n o r m a l subgroup f o r l < i < r .

Then G = M ) n M

x M = 1

x x M

if a n d

( M M - - - M x 2 k

for all subscripts k with 1 < k < r . P r o o f . Suppose first that G = M x G( M M

X 2 x

x M

x x M . Let
r X k

M - ) D M
k x 2

for some subscript k, and write x = n n - - U k - i , w i t h n , G M . W e can then w r i t e x = m m m , where r r n = n for 1 < i < k, and m ; = 1 for i > k, a n d we can get a second decomposition of the element x by t a k i n g m = 1 for a l l subscripts i different from k a n d m = x . In b o t h situations, m G M , for 1 < i < r , a n d thus since we are w o r k i n g i n a direct product, the factors m are uniquely determined. I n the first case, we h a d m = 1, and i n the second, m = x , and thus x = 1.
t x 2 r { t k t t k k

Conversely, assume t h a t { M M M _ ) D M = 1 for 1 < k < r , and suppose t h a t m m m = n n n , where m n G for 1 < i < r. We want m , = n for a l l i, a n d so we suppose t h a t is not true, and we let k be the largest subscript for w h i c h m ^ n . T h e n m m m = n n n , and i n particular, k > 1. W e have
X 2 k x k x 2 r x 2 r { k k x 2 k x 2 k

(n n ---n - )- {m m ---m - )
x 2 k 1 x 2 k x

= {n ){m )k k X

,
k x

and

this element lies i n M

k k

X 1

n M

since M

M _

is a group.

It follows t h a t { n ) { m ) -

= 1, w h i c h is a c o n t r a d i c t i o n .
x 2 r

In particular, if G = M x M x x M is an i n t e r n a l direct product, then M j D M = 1 i f j ^ fc. If r > 2, however, this necessary c o n d i t i o n is definitely not sufficient i n the s i t u a t i o n of T h e o r e m X . 2 2 .
f c

We group of N . [m,n] [ , ] which

m n

need an easy general fact: if M a n d N are n o r m a l subgroups of a G , a n d M n N = 1, then all elements of M commute w i t h a l l elements T o see this, let m G M and n G N , and consider the c o m m u t a t o r of m and n , which, by definition, is the element m ^ n ^ m n . Then = - m , w h i c h lies i n M because M < G . A l s o , [m,n] = (n" )""/!, lies i n A^ because N < G. T h u s [m, n] = 1, a n d this yields m n = n m .
l n 1 m 2

If G = M i x M

x x M
3

is an i n t e r n a l direct product, we conclude

k

t h a t the elements of M

commute w i t h the elements of M

if j ^ k, a n d

344

Appendix:

The

Basics

it follows easily t h a t G is the direct p r o d u c t of the subgroups M i i n every possible ordering of these subgroups. G i v e n a finite collection X of n o r m a l subgroups of a group G , therefore, it is meaningful to say t h a t G is the direct p r o d u c t of the members of X w i t h o u t specifying an ordering. We have already seen t h a t the external direct p r o d u c t P of groups G is the internal direct p r o d u c t of certain subgroups N i < P , where N ^ G . A n o t h e r connection between external and internal direct products is the following.
{ t {

X . 2 3 . L e m m a . L e t G be a g r o u p , a n d suppose that G is the i n t e r n a l direct p r o d u c t of t h e n o r m a l s u b g r o u p s M f o r i < i < r . T h e n G i s i s o m o r p h i c t o t h e e x t e r n a l d i r e c t p r o d u c t of t h e g r o u p s M i .

t

P r o o f . L e t P be the external direct product of the M m a p 9 from P to G by setting 9 ( ( m i , m 2 , , m ) ) = m\m

r 2

a n d construct a

T h e n 9 is a bijection since every element of G is uniquely of the form mim ---m with ra,eM, T o show that 9 is a n i s o m o r p h i s m , we need
2 r

(xix for x , y e M ,
i l l

x )(yiy
r

y ) = x yix y
r x 2 X i

x y
r

and this is clear since

and

V j

commute w h e n i > j . U

For an example of how this can be used, suppose t h a t G is an internal direct p r o d u c t of n o r m a l subgroups, each of w h i c h has a t r i v i a l center. T o prove t h a t Z ( G ) = 1, it suffices to prove t h a t Z ( P ) = 1, where P is the isomorphic group constructed as the external direct p r o d u c t of the factors of G . It is clear, however, t h a t a n r-tuple ( m , m , . . . , m ) is central i n P if a n d only i f each component is central i n its respective group, and it follows t h a t Z ( P ) = l .
x 2 r

Index

abelian Hall subgroup and 7r-length, 94 abelian normal subgroups all cyclic, 189 abelian Sylow subgroup intersection, 38 action, 1 action on cosets, 3, 148 action on subspaces of vector space, 228jf action on right transversal, 148 action via automorphisms, 68 action, Frobenius, 177ff action, semiregular, 177 actions fixing prime order elements, Ulff additive commutator in ring, 125 adjoint (classical) of matrix, 321 admit (of action), 131 Alperin, J., 55, 323 alternating group, 34 alternating group, definition, 326 alternating group, Schur multiplier of, 152 alternating group, simplicity of, 250 arrows in transitive action, 262 augmentation ideal, 315 augmentation map, 314 automorphism, definition, 334 automorphism of symmetric group, 33, 250 automorphism tower, 278 automorphism, order of, 70 axiom of choice, 333 Baer theorem, 55 Baer trick, 142 bar convention, 23 Bartels theorem, 290 base group of wreath product, 73 basis for free abelian group, 315 Bender, H., 31, 217, 271 Berkovich, Y., 86

binomial coefficient, 9 block (of action), 237 Bochert theorem, 246^ Brodkey theorem, 38, 71, 74 Burnside, book of, 181, 217 Burnside orbit count, 7 Burnside pV-theorem, 216# Burnside p-nilpotence theorem, 159 Burnside theorem, prime degree action, 240 Burnside, W., 30 Cameron, P., 249, 285 canonical homomorphism, 338 Carter subgroup, 91 Cauchy theorem, 8, 10 Cauchy-Frobenius orbit count, 7 center of group, 2, 334 central extension, 151 central prime order elements, 174 central series, 20, 115 central subgroup, transfer to, 154 centralizer, definition, 339 Cermak-Delgado subgroup, 43 chain stabilizer, 137 character theory, 53, 59, 147, 182, 216-217 character, permutation, 236 characteristic series, 7r-separable, 92 characteristic subgroup, 11 characteristic subgroup, definition, 335 characteristically simple group, 277 Chermak-Delgado theorem, 41 class 2 group, 122, 142 class (conjugacy), 4 class equation, 5 class (nilpotence), 22, 116, 119 class size, 6

345

346

Index

class sizes, common divisor graph, 268 class sizes, number of, 128 class sizes, smallest, 128 classical adjoint of matrix, 321 collection of commutators, 120ff common divisor graph on class sizes, 268 common-divisor graph, components, 265ff commutator, 40, 48, 113jff commutator, additive, 125 commutator collection, I20ff commutator, multiple, 115 commutator subgroup, 80 see also derived subgroup commutators in coprime action, 138 complement, 85 complement for normal subgroup, 65 complete group, 279 component of group, 273ff, 287 see also layer components of common-divisor graph, 265# composition factor, 29 composition series, 29 composition series of solvable group, 84 conjugacy class, 4 see also class conjugacy of complements, 82 conjugacy of Sylow subgroups, 14 conjugates of subgroup, 6 conjugation action, 2 conjugation action on normal subgroup, 230 connected graph, 260 control of fusion, 158, 167 control of transfer, 167, 295# coprime action, 96# coprime action on abelian group, 140 coprime action, orbit sizes of, 102ff coprime action, commutators in, 138 coprime orbit sizes, 102-105 core of cyclic subgroup, 63 core, 3 corefree subgroup, 285 correspondence theorem, 340 coset, 331 coset action, 3 coset, double, 6, 304 crossed homomorphism, 76, 114 cyclic extension, 107 cyclic factor group, 118 cyclic group automorphism, 334 cyclic group, subgroups of, 330 cyclic subgroup, core of, 63 cyclic Sylow subgroups, 159# Dedekind lemma, 328 degree, 225 depth, subnormal, 45

derived length, 82 derived length and nilpotence class, 128 derived length, large, 146 derived series, 80 derived subgroup, 80, 113 derived subgroup,finiteness,155 derived subgroup, noncommutators in, 125 development, 15 Dietzmann theorem, 156 dihedral group, 55ff, 120, 189, 196 dihedral group, construction, 70 direct diamond, 329 direct product, 69# direct product, center of, 25 direct product, definition, 342 Doerk, K., 86 dot action on right transversal, 148 double coset, 6, 304 double transitivity, 225 E, 274 see also layer elementary abelian group, 27, 82, 202, 298, 310 Euler totient function, ip, 334 even permutation, 34, 326 existence of Sylow subgroup, 8, 9, 12 exponent, 116 extension, 66 external direct product, 342 extraspecial p-group, 123 F, 25 see also Fitting subgroup F*, 271 see also generalized Fitting subgroup factor group, 337 faithful action, 2, 40, 133, 233 faithful action on normal subgroup, 145 faithful orbit (existence of), 74 Feit-Thompson theorem, 30, 75, 97 Feit, W., 30 finite index center, Schur theorem, 155 Fischer-Greiss monster, 31 Fitting subgroup, 25, 46, 53, 271 Fitting subgroup and subnormality, 47 Fitting subgroup of solvable group, 86 Fitting theorem, coprime action, 140 fixed points come from fixed points, 101 fixed-point subgroup, 96 focal subgroup theorem, 165 Frattini argument, 15 Frattini subgroup, 27, 95, 282, 330 Frattini subgroup of p-group, 27, 117 free abelian group, 315 Frobenius action, 177^, 232 Frobenius complement, 26, 179 Frobenius complement, nonsolvable, 181 Frobenius complement, odd order, 185, 193

Index

347

Frobenius complement, properties, 186^, 193 Frobenius group, 181 Frobenius kernel, 179 Frobenius kernel, nilpotence of, 196#, 201 Frobenius kernel, nonabelian, 180 Frobenius kernel, solvable, 196 Frobenius p-nilpotence theorem, 171jff, 197 Frobenius theorem on kernel existence, 181, 186 Frobenius theorem, permutations, 183 Frobenius, G., 216 fundamental counting principle, 5 Furtwangler, P., 313 fused classes, 100, 158 fusion and p-nilpotence, 170 fusion, control of, 158 general linear group, 30, 203#, 228 general linear group, order of, 204 general linear gronp, Sylow subgroup of, 205 generalized Brodkey theorem, 39 generalized dihedral group, 57, 70 generalized Fitting subgroup, 271ff,276, 281 generalized quaternion group, 74, 110, 120, 153, 189, 196, 209 generating set of subgroup, 329 glasses, 96 GL, 30 see also general linear group Glauberman Z( J)-theorem, 217 Glauberman lemma, 97 Glauberman, G., 31, 217 global property, 59 Goldschmidt, D., 31, 55, 217 good theorem, 2 graph and transpositions, 241 graph, common divisor, 265ff graphs and primitivity, 260 graphs and orbitals, 259 greed, 46 group, definition, 325 group of order 120, 18, 256 group of order 21952, 17 group of order 24, 33 group of order 8, 190 group of order p q , 37 group of order p q , 32 group of order p g , 33 group of order p q , 37 group of order p q , 32 group of order p g r , 38 group ring, 313 groups of small order, 38
2 2 2 3 a

Hall C-theorem, 87 Hall D-theorem, 90 Hall E-theorem, 86 Hall E-theorem converse, 87 Hall 7r-subgroup, 86 Hall w-subgroups, nonisomorphic, 90 Hall subgroup, 12 Hall subgroup in solvable group, 86 Hall subgroup in 7r-separable group, 93 Hall subgroup, conjugacy, 87 Hall subgroup, normal, 75 Hall-Higman Lemma 1.2.3, 93 Hall-Wielandt transfer theorem, 167, 297 Hall-Witt identity, 125 Hall, P., 86, 120, 134 Hartley-Turull theorem, 102 Hawkes, T., 86 Higman, G., 196 Higman-Sims group, 257 Hochschild, G., 313 homomorphism, definition, 337 homomorphism theorem, 339 homomorphism, canonical, 338 Horosevskii theorem, 70 Huppert theorem on metacyclic Sylows, 308 imprimitive action, 237 induced action on factor group, 132 infinite p-group, 8, 19 inner automorphism (definition), 335 internal direct product, 342 intersection of abelian subgroups, 61 intersection of subnormal subgroups, 47 intersection of Sylow subgroups, 37-39 invariant classes, 100 invariant cosets, 101 invariant Sylow subgroups, 96 involution, 55 involutions, nonconjugate, 57 Ishikawa, K., 128 isomorphism, 333 Iwasawa lemma, 252 J , 198 see also Thompson subgroup Jacobi identity, 125 Janko group, 164 join of subgroups, 42 join of subnormal subgroups, 47ff Jordan set, 243 Jordan theorem, 211ff Jordan-Holder theorem, 29, 84 Jordan, C., 227 Kegel, O., 294 kernel of homomorphism, definition, 338 kernel of action, 2 kernel of crossed homomorphism, 76

half transitivity, 232jf

348

Index

kernel of transfer, 165 Klein group, 34, 46, 56 Kuo, T., 323 Lagrange theorem, 331 Lagrange theorem, converse, 9, 24 lattice, 42, 47 layer, 274ff, 282, 287 left transversal, 77 Lie ring, 126 linear representation, 147 local subgroup, 58 local subgroup and homomorphism, 59-60 local-to-global theorem, 59 lower central series, 116, 127, 281 lower triangular matrix, 205 Lucchini theorem, 63 Lyons, ft, 55 Mackey transfer theorem, 304 magic eyeglasses, 96 Mann subgroup, 128 Manning theorem and 3-transitivity, 264 Maschke theorem, 309 Mathieu group, 31, 226#, 256# matrix unit, 253 Matsuyama, H., 31, 55, 217 maximal class p-group, 120 maximal subgroup, 25, 27 maximal subgroup, definition, 329 maximal subgroup of solvable group, 84, 91 maximal subgroup, nilpotent, 168, 209 McKay, J., 8 measure, Chermak-Delgado, 41ff metacyclic group, 160, 308^ minimal-normal subgroup, 48, 54, 82, 90, 275 minimal-simple group, 61 monster simple group, 31 multiple commutator, 115 multiple transitivity, 225, 264 multiple transitivity and primitivity, 240 N-group, 61 n! theorem, 4 Navarro, G., 164 Neumann, P., 7 nilpotence and subnormality, 47 nilpotence class, 22, 116, 119, 296 nilpotence class and derived length, 128 nilpotence class of Mann subgroup, 129 nilpotent factor group, 27 nilpotent group, 20, 24 nilpotent injector, 91 nilpotent joins, Baer theorem, 55 nilpotent maximal subgroup, 168, 209 nilpotent normal subgroup, 26

nonabelian simple group, 29 normal p-complement, 159, 164 normal abelian subgroups cyclic, 189 normal closure, 51 normal Hall subgroup and splitting, 75 normal series, 20, 80 normal subgroup, definition, 336 normal subgroup of primitive group, 238 normal-J theorem, 210 normal-P theorem, 207 normalizer, 3, 336 normalizer of subnormal subgroup, 48 normalizers grow, 22 O'Brien, E., 19 odd permutation, 34, 326 odd-order theorem, 30 see also Feit-Thompson theorem f i r of p-group, 120 Hi of class 2 p-group, 122 orbit, 4, 223 orbit size, 5 orbits, number of, 7 orbit sizes in coprime actions, 102 orbital, 257 see also suborbit orbits of subgroup, 236 order of automorphism, 70 P x Q-theorem, 139 P x Q-theorem, strong form, 144 p V theorem, 30, 217^ p-central element, 220 p-complement, 85, 88 p-complement, normal, 159 p-cycle in primitive group, 245 p-group, 8 p-group of class 2, 122 p-group of maximal class, 120 p-group, center of, 20 p-group, elementary abelian, 27 see also elementary abelian p-group, infinite, 8 p-group, nilpotence of, 21 p-group, omega subgroup of, 120 p-group, subgroups of, 24 p-group with unique minimal subgroup, 189 p-groups, number of, 19 p-local subgroup, 58 p-local subgroup of p-solvable group, 140 p-local subgroups and p-nilpotence, 171 p-nilpotence and fusion, 170 p-nilpotent group, 159 see also normal p-complement p-solvable group, 93, 95, 107, 140, 146, 207, 210 paired orbital, 257, 262 partitioned group, 186, 195, 232

Index

349

Passman, D., 232, 240 path connected graph, 260 perfect group, 151 permutable subgroups, 6, 49 permutation character, 7, 236 permutation group, 223ff permutation isomorphism, 5, 103, 224 permutation representation, 2, 223 Pettet, M., 278 27 see also Frattini subgroup PGL, 205 7r-group, 12 7r-length, 94 7r-separable group, 91 7r-solvable group, 93 point stabilizer, 3 point stabilizer, primitive group, 239 pointwise stabilizer of set, 242 Praeger, C , 106, 249, 265, 285 pretransfer map, 149 prime degree action, Burnside theorem, 240 prime index subgroup, 6, 18, 85 prime order elements central, 174 prime subdegree, 269 primitive action, 237# primitive group containing p-cycle, 241, 242, 245 primitive group, simplicity criterion, 252 primitive solvable group, 248 primitive subgroup of symmetric group, 246# primitivity and multiple transitivity, 240 principal ideal theorem (of transfer), 313 pro-p-group, 18 product of subgroups, 6, 49, 327, 336 projective general linear group, 205 projective special linear group, 30 see also PSL PSL, 30, 205 PSL, simplicity, 251ff quasinormal subgroup, 50 quasiquaternion, 123 quasisimple group, 272 quaternion group, 74 see also generalized quaternion group quotient group, 337 rank of abelian p-group, 203 regular p-group, 297, 307 regular action, 2, 225 regular orbit, existence, 71 regular subgroup of permutation group, 235 regular wreath product, 73 repeated commutator, 132^ representation, 147 representation group (Schur), 151, 272

representation, permutation, 2 right transversal, 77 right transversal, dot action on, 148 Saxl, J., 249, 285 Schenkman theorem, 283 Schenkman, E., 278 Schur multiplier, 123, 151, 153, 272 Schur representation group, 151, 272 Schur theorem, finite index center, 155 Schur-Zassenhaus theorem, 75# second center, 20 Seitz, G., 249, 285 self-centralizing subgroup, 129 self-normalizing Sylow subgroup, 164 semidihedral group, 74, 120, 189, 196 semidihedral group, Schur multiplier, 153 semidirect product, applications of, 70 semidirect product, existence of, 72 semiregular action, 177, 225 semisimple group, 274 series of subgroups, 20 setwise stabilizer of set, 242 sharp multiple transitivity, 226 sharp transitivity, 225 simple group, nonabelian, 29 simple group classification, 30 simple group, order of, 163 simple group, Sylow 2-subgroup of, 162 simple groups of small order, 35 Sims conjecture, 249, 285 SL, 30, 110 SL, order of, 204 socle, 48, 54 solvable group, 30, 80# solvable group, Fitting subbgroup, 86 solvable group, maximal subgroup, 84 solvable minimal normal subgroup, 82 solvable primitive group, 248 special linear group, 30 see also SL split extension, 65^ sporadic simple group, 30 stabilizer of chain, 137 stabilizer of point, 3 strongly conjugate subgroups, 54, 289ff strongly Jordan set, 243 subdegree, 259 subdegrees in imprimitive group, 265 subdegrees in primitive group, 26Iff subnormal 7r-subgroup, 53 subnormal closure, 289ff subnormal core, 294 subnormal depth, 45 subnormal nilpotent subgroup, 47 subnormal partition, 195 subnormal series, 7>separable, 92 subnormal subgroup, 45

350

Index

subnormality, 45ff, 271ff suborbit, 257ff supersolvable group, 85 supersolvable group, Mann subgroup of, 131 Suzuki, M., 55 Suzuki group, 163 Sylow p-subgroup of symmetric group, 299 Sylow C-theorem, 14 Sylow counting theorem, 16 Sylow D-theorem, 15 Sylow E-theorem, 9 Sylow subgroup, 8 Sylow subgroup of Frobenius complement, 188, 193 Sylow subgroup of general linear group, 205 Sylow subgroup, cyclic, 159ff Sylow subgroups, intersection, 10, 16, 38 Sylow subgroups, number of, 15-16 Sylow subgroups, set of, 10, 14 Sylow system, 90 Sylow theorem, via Cauchy, 8, 12 symmetric group, 4, 325 symmetric group automorphism, 250 symmetric group, normal subgroup of, 251 symmetric group, primitive subgroup of, 246 Tate's theorem, 168 Taussky-Todd, O., 196 the, 11, 334 Thompson P x Q-theorem, 139 Thompson p-nilpotence theorem, 197, 201, 213 Thompson subgroup, 198, 20Iff Thompson theorem on Sims conjecture, 285ff Thompson, J., 30, 61, 169, 217 Thompson's thesis, 179, 196, 201, 214 three subgroups lemma, 126 totient function, 334 transfer evaluation lemma, 153ff transfer kernel, 165 transfer map, 149 transfer theory, 147ff, 295ff transfer to central subgroup, 154 transfer to derived subgroup, 313 transitive action, 7, 223 transitivity of transfer, 301 translate of subset in action, 237 transposition, 240, 326 transversal, 77 see also right transversal trivial block, 237 upper central series, 20 upper triangular matrix, 204

Verlagerung, 149 weakly closed subgroup, 163 weight of commutator, 127 Weiss theorem on subdegrees, 262 Wielandt automorphism tower theorem, 278 Wielandt solvability theorem, 88 Wielandt subgroup, 54 Wielandt zipper lemma, 50ff wreath product, 73, 124, 296 Yoshida theorem, 296 Yoshida, T., 167 Zassenhaus, H., 181 Zenkov theorem, 61 zipper lemma, 50

Titles in This Series

97 D a v i d C . U l l r i c h , Complex made simple, 2008 96 N . V . K r y l o v , Lectures on elliptic and parabolic equations in Sobolev spaces, 2008 95 L e o n A . T a k h t a j a n , Quantum mechanics for mathematicians, 2008 94 James E . H u m p h r e y s , Representations of semisimple Lie algebras in the BGG category O, 2008 93 Peter W . M i c h o r , Topics in differential geometry, 2008 92 I. M a r t i n Isaacs, Finite group theory, 2008 91 Louis Halle Rowen, Graduate algebra: Noncommutative view, 2008 90 L a r r y J . Gerstein, Basic quadratic forms, 2008 89 A n t h o n y B o n a t o , A course on the web graph, 2008
88 N a t h a n i a l P . B r o w n a n d N a r u t a k a O z a w a , C*-algebras and finite-dimensional

approximations, 2008
87 S r i k a n t h B . Iyengar, G r a h a m J . Leuschke, A n t o n L e y k i n , C l a u d i a M i l l e r , E z r a

M i l l e r , A n u r a g K . Singh, a n d U l i W a l t h e r , Twenty-four hours of local cohomology, 2007 86 Y u l i j Ilyashenko a n d Sergei Yakovenko, Lectures on analytic differential equations, 2007
85 J o h n M . A l o n g i a n d G a i l S. N e l s o n , Recurrence and topology, 2007 84 C h a r a l a m b o s D . A l i p r a n t i s a n d R a b e e T o u r k y , Cones and duality, 2007

83 Wolfgang E b e l i n g , Functions of several complex variables and their singularities (translated by Philip G. Spain), 2007
82 Serge A l i n h a c a n d P a t r i c k G r a r d , Pseudo-differential operators and the Nash-Moser

theorem (translated by Stephen S. Wilson), 2007 81 V . V . Prasolov, Elements of homology theory, 2007 80 Davar Khoshnevisan, Probability, 2007 79 W i l l i a m Stein, Modular forms, a computational approach (with an appendix by Paul E. Gunnells), 2007 78 H a r r y D y m , Linear algebra in action, 2007 77 Bennett C h o w , P e n g L u , a n d L e i N i , Hamilton's Ricci flow, 2006 76 M i c h a e l E . Taylor, Measure theory and integration, 2006 75 Peter D . M i l l e r , Applied asymptotic analysis, 2006 74 V . V . Prasolov, Elements of combinatorial and differential topology, 2006 73 Louis H a l l e Rowen, Graduate algebra: Commutative view, 2006 72 R . J . W i l l i a m s , Introduction the the mathematics offinance,2006 71 S. P . Novikov a n d I. A . Taimanov, Modern geometric structures andfields,2006 70 Sean Dineen, Probability theory infinance,2005 69 Sebastian M o n t i e l a n d A n t o n i o R o s , Curves and surfaces, 2005 68 Luis Caffarelli a n d Sandro Salsa, A geometric approach to free boundary problems, 2005 67 T . Y . L a m , Introduction to quadratic forms over fields, 2004
66 Y u l i E i d e l m a n , V i t a l i M i l m a n , a n d A n t o n i s T s o l o m i t i s , Functional analysis, An

introduction, 2004 65 S. R a m a n a n , Global calculus, 2004 64 A . A . K i r i l l o v , Lectures on the orbit method, 2004 63 Steven Dale C u t k o s k y , Resolution of singularities, 2004 62 T . W . K r n e r , A companion ;to analysis: A second first andfirstsecond course in analysis, 2004

TITLES IN THIS SERIES

61 T h o m a s A . Ivey a n d J . M . L a n d s b e r g , Cartan for beginners: Differential geometry via moving frames and exterior differential systems, 2003
60 A l b e r t o C a n d e l a n d L a w r e n c e C o n l o n , Foliations II, 2003

59 Steven H . W e i n t r a u b , Representation theory of finite groups: algebra and arithmetic, 2003 58 C e d r i c V i l l a n i , Topics in optimal transportation, 2003 57 R o b e r t P l a t o , Concise numerical mathematics, 2003 56 E . B . V i n b e r g , A course in algebra, 2003 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34
33

C . H e r b e r t Clemens, A scrapbook of complex curve theory, second edition, 2003 A l e x a n d e r B a r v i n o k , A course in convexity, 2002 H e n r y k Iwaniec, Spectral methods of automorphic forms, 2002 Ilka A g r i c o l a a n d T h o m a s Friedrich, Global analysis: Differential forms in analysis, geometry and physics, 2002 Y . A . A b r a m o v i c h a n d C . D . A l i p r a n t i s , Problems in operator theory, 2002 Y . A . A b r a m o v i c h a n d C . D . A l i p r a n t i s , An invitation to operator theory, 2002 J o h n R . H a r p e r , Secondary cohomology operations, 2002 Y . Eliashberg a n d N . M i s h a c h e v , Introduction to the ^-principle, 2002 A . Y u . K i t a e v , A . H . Shen, a n d M . N . V y a l y i , Classical and quantum computation, 2002 Joseph L . Taylor, Several complex variables with connections to algebraic geometry and Lie groups, 2002 Inder K . R a n a , An introduction to measure and integration, second edition, 2002 J i m A g l e r a n d J o h n E . M c C a r t h y , Pick interpolation and Hilbert function spaces, 2002 N . V . K r y l o v , Introduction to the theory of random processes, 2002 J i n H o n g a n d Seok-Jin K a n g , Introduction to quantum groups and crystal bases, 2002 G e o r g i V . Smirnov, Introduction to the theory of differential inclusions, 2002 R o b e r t E . Greene a n d Steven G . K r a n t z , Function theory of one complex variable, third edition, 2006 L a r r y C . G r o v e , Classical groups and geometric algebra, 2002 E l t o n P . H s u , Stochastic analysis on manifolds, 2002 H e r s h e l M . Farkas a n d Irwin K r a , Theta constants, Riemann surfaces and the modular group, 2001 M a r t i n Schechter, Principles of functional analysis, second edition, 2002 James F . Davis a n d P a u l K i r k , Lecture notes in algebraic topology, 2001 Sigurdur Helgason, Differential geometry, Lie groups, and symmetric spaces, 2001
D m i t r i B u r a g o , Y u r i B u r a g o , a n d Sergei Ivanov, A course in metric geometry, 2001

32 R o b e r t G . B a r t l e , A modern theory of integration, 2001 31 R a l f K o r n a n d E l k e K o r n , Option pricing and portfolio optimization: Modern methods of financial mathematics, 2001 30 J . C . M c C o n n e l l a n d J . C . R o b s o n , Noncommutative Noetherian rings, 2001 29 Javier Duoandikoetxea, Fourier analysis, 2001 28 L i v i u I. Nicolaescu, Notes on Seiberg-Witten theory, 2000 27 T h i e r r y A u b i n , A course in differential geometry, 2001 26 R o l f B e r n d t , An introduction to symplectic geometry, 2001

F o r a c o m p l e t e list o f t i t l e s i n t h i s series, v i s i t t h e

A M S B o o k s t o r e a t www.ams.org/bookstore/.

The text begins with a review of group actions and Sylow theory. It includes semidlrect products, the Schur-Zassenhaus theorem, the theory of commutators, coprime actions on groups, transfer theory. Frobenius groups, primitive and multiply transitive permutation groups, the simplicity of the PSL groups, the generalized Fitting subgroup and also Thompsons J-subgroup and his normal p-complement theorem. Topics that seldom (or never) appear in books are also covered These include subnormality theory, a group-theoretic proof of Burnside's theorem about groups with order divisible by just two primes, the Wielandt automorphism tower theorem. Yoshida's transfer theorem, the "principal ideal theorem" of transfer theory and many smaller results that are not very well known. Proofs often contain original ideas, and they are given in complete detail. In many cases they are simpler than can be found elsewhere.The book is largely based on the author's lectures, and consequently, the style is friendly and somewhat informal. Finally, the book includes a large collection of problems at disparate levels of difficulty.These should enable students to practice group theory and not just read about ft Martin Isaacs is professor of mathematics at the University of Wisconsin. Madison. Over the years, he has received many teaching awards and is well known for his inspiring teaching and lecturing. He received the University of Wisconsin Distinguished Teaching Award in 1985. the Benjamin Smith Reynolds Teaching Award in 1989, and the Wisconsin Section MAA Teaching Award in 1993. to name only a few. He was also honored by being the selected MAA Plya Lecturer in 2003-2005.

ISBN 9/8-0-8218 *3**-*

For additional information and updates on this book, visit

www.ams.org/bookpages/gsm-92

9 780821 8*)344<.
A . M S nu tl- WVh

www.ams.org